Copyright TechTarget - All rights reserved https://cyber.law.harvard.edu/rss/rss.html Techtarget Feed Generator en Tue, 21 Jul 2026 05:51:31 GMT https://www.techtarget.com/searchnetworking editor@techtarget.com <p>Cloud adoption, hybrid work, SaaS, IoT and distributed networks are making enterprise networks more complex than ever. Traditional network management -- based on manual configuration, monitoring and human intervention after issues occur -- cannot keep pace.</p> <p>Enter AI-driven intelligent network automation. The technique has emerged as a tool for business continuity, digital transformation and network resilience. Among other capabilities, it enables prediction, automation and self-healing -- positioning modern networks as strategic differentiators.</p> <p>This article traces the evolution of predictive networks and examines how enterprises can combine AI-driven observability, intent-based automated networking and governance to enable intelligent network automation and support business growth.</p> <section class="section main-article-chapter" data-menu-title="The shift from reactive to predictive networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The shift from reactive to predictive networking</h2> <p>Traditional network operations are hobbled by built-in limitations that cannot meet today's requirements. These shortcomings include limited visibility across diverse platforms and reactive workflows that depend on manual troubleshooting and post-incident remediation.</p> <p>Network operations teams suffer from alert fatigue as they try to comb through operational data that grows across logs, traces and other monitoring sources. Together, these factors lead to slow <a href="https://www.techtarget.com/searchDisasterRecovery/feature/Disaster-recovery-glossary-Terms-to-know">mean time to detect (MTTD) and mean time to resolve (MTTR) metrics</a>, fueling service-level agreement issues or significant penalties.</p> <p>These limitations are no longer sustainable as businesses rely more on digital services and users expect improved <a href="https://www.techtarget.com/searchITOperations/feature/How-digital-employee-experience-strategy-drives-productivity">digital employee experience</a>. Applications require higher performance and availability, while security requirements are becoming even stricter.</p> <p>Intelligent network automation is the next step in the evolution of network operations. It shifts focus from issue response to issue prevention, with the network itself automatically maintaining service quality.</p> </section> <section class="section main-article-chapter" data-menu-title="AI-driven observability and autonomous operations"> <h2 class="section-title"><i class="icon" data-icon="1"></i>AI-driven observability and autonomous operations</h2> <p>Building an effective AI-driven operations foundation <a href="https://www.techtarget.com/searchnetworking/tip/Network-observability-vs-monitoring-Whats-the-difference">begins with observability</a> and data quality. AI operations require comprehensive, high-quality telemetry across on-premises, cloud and remote platforms. This includes networks, applications, cloud services, endpoints and security tools. Unified observability and quality data provide the visibility needed for <a href="https://www.techtarget.com/searchbusinessanalytics/definition/predictive-analytics">predictive analytics</a>.</p> <p>Specific technologies enable predictive analytics for network operations, including the following:</p> <ul class="default-list"> <li><b>AIOps platforms. </b>Correlate and analyze operational data.</li> <li><b>Machine learning models.</b> Identify patterns and predict failures.</li> <li><b>Agentic AI systems.</b> <a href="https://www.techtarget.com/searchenterpriseai/definition/agentic-AI">Capable of autonomous decision-making</a> and execution.</li> </ul> <p>Autonomous remediation establishes self-healing workflows that fix common issues without human intervention and before they affect users. It also includes automated root-cause analysis capabilities to help avoid future incidents.</p> </section> <section class="section main-article-chapter" data-menu-title="Intent-based networking and operational transformation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Intent-based networking and operational transformation</h2> <p>Transforming network operations to a predictive, AI-driven approach starts with <a href="https://www.techtarget.com/searchnetworking/infographic/What-is-intent-based-networking-and-how-does-it-work">intent-based networking</a>. IBN turns business objectives into automated network policies, aligning strategic goals with network operations tactics. It shifts operations away from device-by-device configuration to a policy-driven approach.</p> <p>IBN capabilities include the following:</p> <ul class="default-list"> <li>Automated policy deployment and enforcement.</li> <li>Continuous validation that network behavior matches business intent.</li> <li>Dynamic adaptation to changing application, user and business requirements.</li> </ul> <p>IBN supports autonomous networking by providing a policy framework that guides AI-driven actions. This approach reduces configuration errors, smooths operational inconsistencies, improves governance and enables compliance. Each of these results supports network operations that power organizational agility.</p> </section> <section class="section main-article-chapter" data-menu-title="The evolving role of NetOps teams"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The evolving role of NetOps teams</h2> <p>It's not just the network that evolves; network operations teams transform as well. <a href="https://www.techtarget.com/searchnetworking/tip/How-NetOps-differs-from-traditional-network-operations">NetOps teams</a> are shifting from network operators to AI supervisors. Teams spend less time on manual configuration tasks and repetitive troubleshooting, allowing them to focus on innovation.</p> <p>New responsibilities include the following:</p> <ul class="default-list"> <li>Supervise AI-driven recommendations and automated actions.</li> <li>Establish governance policies and operational guardrails.</li> <li>Manage automation strategies and exception handling.</li> <li>Collaborate with cloud, security and business teams.</li> </ul> <p>These responsibilities drive an evolution of skills that emphasizes automation, analytics and AI governance.</p> <p>Most enterprises begin with AI-assisted recommendations before enabling autonomous remediation. Human oversight remains vital for critical decisions, governance and accountability. Use an effective <a href="https://www.techtarget.com/searchDataCenter/tip/AI-operating-models-Balancing-autonomy-and-human-oversight">AI operating model</a> that balances oversight with autonomy.</p> </section> <section class="section main-article-chapter" data-menu-title="The path to intelligent network automation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The path to intelligent network automation</h2> <p>Intelligent network automation does not mean companies have to replace their current networks. However, they require new approaches, modern governance and improved data collection. In many cases, organizations are further along in their transformation than they realize.</p> <p>Evaluate the network's current status and establish a phased modernization approach with the following progression:</p> <ol class="default-list"> <li>Recognize traditional monitoring and alerting.</li> <li>Integrate enhanced observability.</li> <li>Create AIOps-driven analytics and recommendations.</li> <li>Implement automated remediation.</li> <li>Transition to fully predictive and autonomous operations.</li> </ol> <p>Prioritize high-value use cases to demonstrate the validity and significance of intelligent networks. As automation expands, ensure governance, security and data quality mature alongside it.</p> </section> <section class="section main-article-chapter" data-menu-title="Business value and ROI of intelligent network automation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Business value and ROI of intelligent network automation</h2> <p>Intelligent network automation supports digital transformation, scalability and future growth. Evaluate its significance in operational and financial terms.</p> <p>Operational benefits include the following:</p> <ul class="default-list"> <li>Improved uptime and service availability.</li> <li>Faster incident detection and remediation.</li> <li>Improved operational resilience and business continuity.</li> <li>Reduced alert fatigue and support staff frustration.</li> <li>Increased opportunity for network teams to focus on innovation.</li> </ul> <p>Financial benefits include the following:</p> <ul class="default-list"> <li>Reduced operational overhead.</li> <li>Lower costs and penalties associated with <a target="_blank" href="https://www.erwoodgroup.com/blog/the-true-costs-of-downtime-in-2025-a-deep-dive-by-business-size-and-industry/" rel="noopener">outages and downtime</a>.</li> <li>More efficient use of IT resources.</li> </ul> <p>Essential KPIs and metrics for measuring success include the following:</p> <ul class="default-list"> <li>Reduced MTTD and MTTR.</li> <li>Increased network availability and performance.</li> <li>Higher automation rates.</li> <li>Lower incident volumes and operational costs.</li> <li>Improved <a href="https://www.techtarget.com/searchitoperations/tip/How-DEX-metrics-help-build-a-better-digital-workplace">digital employee experience scores</a>.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Make the shift to intelligent network automation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Make the shift to intelligent network automation</h2> <p>Intelligent network automation shifts network management from infrastructure control to orchestrating intelligent, self-optimizing network operations. Organizations that deploy intelligent network automation -- combining AI-driven observability, intent-based automation, strong governance and human oversight -- improve resilience, reduce risk and support future business demands.</p> <p><i>Damon Garn owns Cogspinner Coaction and provides freelance IT writing and editing services. He has written multiple CompTIA study guides, including the Linux+, Cloud Essentials+ and Server+ guides, and contributes extensively to InformaTechTarget, The New Stack and CompTIA Blogs.</i></p> </section> Intelligent technologies that include predictive analytics and intent-based networking are reshaping how enterprises oversee their complex corporate networks. https://cdn.ttgtmedia.com/rms/onlineimages/machine_learning_g1307219089.jpg https://www.techtarget.com/searchnetworking/tip/How-intelligent-network-automation-transforms-enterprise-operations Fri, 17 Jul 2026 12:00:00 GMT <p>The goal of every digital transformation initiative is to drive leaner and easier business operations, create better UX and enable faster decision-making. Every transformation project relies on networks. However, many organizations still operate without a complete view of their entire network stack.</p> <p>According to<a href="https://backlinko.com/digital-transformation-stats"> </a>IDC's Worldwide Digital Transformation Spending Guide, organizations worldwide will <a target="_blank" href="https://my.idc.com/getdoc.jsp?containerId=IDC_P32575" rel="noopener">spend almost $4 trillion</a> on digital transformation by 2027. Despite this capital commitment, the gap between digital transformation's promises and its operational realities keeps widening. <a href="https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/successful-transformations">Research</a> from McKinsey shows that 70% of transformation initiatives fail to meet their objectives.</p> <p>This isn't because companies aren't investing enough in modern cloud deployments, AI tooling or advanced applications. Rather, companies don't fully understand what runs on their networks.</p> <p>As networks become more distributed and sophisticated, it becomes harder to see who and what devices are on the network and detect when a breach occurs. This limited network visibility translates into invisible risks and blind spots that can surface at unexpected times.</p> <section class="section main-article-chapter" data-menu-title="The role of visibility in modern networks"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The role of visibility in modern networks</h2> <p>Digital transformation reshapes how companies rely on their networks. Networks no longer have clear perimeters. In the past, companies had the visibility they needed. They learned how traffic flowed by monitoring their on-premises data centers and relying on a small number of fixed observation points.</p> <p>Now that enterprises are moving to the cloud, apps and services mostly run in a distributed environment. Areas of operation typically include Kubernetes clusters across private on-premises infrastructure, multiple public cloud providers and software-defined WAN edge locations.</p> <p>Hybrid work has further expanded this framework. In just a single second, a transaction can traverse different environments that could be owned and monitored by separate IT teams. This adds complexity to the overall network security posture, making true <a href="https://www.techtarget.com/searchnetworking/tip/End-to-end-network-observability-for-AI-workloads">end-to-end monitoring</a> difficult. The consequences directly undermine digital transformation goals.</p> <p>Some of the most damaging network visibility issues enterprises face today include the following:</p> <h3>1. Application performance declines</h3> <p>Applications that power businesses are only as reliable as the networks used to deliver them. When network visibility is fragmented, IT teams struggle to understand how applications perform in real time. Applications could experience unique slowdowns that go undetected until an end-user complaint signals the issue.</p> <p>Finding the root causes of these performance degradations can be difficult due to the multiple connected paths in the system. This delays time-to-market, strains resources and negatively affects revenue generation.</p> <h3>2. Security blind spots</h3> <p>Transformation inherently grows the attack surface. New tools introduce new entry points and the potential of malicious payloads. But security teams can't protect what they can't observe. Limited network visibility opens the door to attackers who can move laterally across environments undetected. Companies need to roll out comprehensive <a href="https://www.techtarget.com/searchsecurity/tip/Network-visibility-and-monitoring-tools-now-amp-up-security">network visibility and monitoring tools</a> to gain insight into traffic, establish baselines, detect malicious actors and identify identity theft in a timely manner.<b> </b></p> <p>A breach could halt a digital transformation plan entirely. When IT teams redirect engineering efforts toward remediations, it stalls the overall roadmap. Undetected breaches dramatically increase regulatory exposure, remediation costs and reputational damage. Network visibility assurance is an integral part of network security.</p> <h3>2. UX issues</h3> <p>Users often don't know the difference between a <a href="https://www.techtarget.com/searchnetworking/answer/What-key-network-performance-issues-spark-service-disruptions">network problem</a> and an app bug because they both produce the same result: unusable software. Network visibility helps the team identify which layer is responsible.<b> </b>In a distributed network, issues span multiple segments. A network problem could be anything from packet loss at a cloud provider's egress point to elevated latency in a critical region.</p> <p>Without complete traffic data across all segments, problems like these are impossible to triage.<b> </b>User complaints are the first signal of network performance degradation, which can damage trust and lead to productivity losses.</p> <h3>4. Major decisions made based on assumptions</h3> <p>IT leaders can't make network investment decisions without comprehensive network performance data. Otherwise, management is forced to make educated guesses, which could lead to over-provisioning the wrong bottlenecks to compensate for uncertainty. Decisions can also get delayed until after problems become acute.<b> </b>Mistakes made early on often become expensive corrections later.</p> <h3>5. Increased troubleshooting complexity</h3> <p>Fragmented visibility prevents application and security teams from having a full view of performance or security. Even seemingly straightforward troubleshooting can consume significant time and resources. Network teams end up spending more time diagnosing network issues when the root cause is a lack of telemetry.</p> <h3>6. Cloud and hybrid initiatives lose momentum</h3> <p>Cloud migration, <a href="https://www.techtarget.com/searchsecurity/tip/Multi-cloud-security-challenges-and-best-practices">multi-cloud adoption</a> and hybrid networking underpin most enterprise digital transformation strategies. Yet each introduces visibility challenges unlike anything organizations have had to manage before. Left unattended, it can stall digital transformation.</p> </section> <section class="section main-article-chapter" data-menu-title="How to improve network visibility for digital transformation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to improve network visibility for digital transformation</h2> <p>Network teams should ensure they have adequate visibility before undertaking a digital transformation project. Ways to improve network visibility include the following:</p> <ol type="1" start="1" class="default-list"> <li><b>Adopt a unified observability app for hybrid and multi-cloud environments. </b>While monitoring identifies the failures, observability shows how it failed. Modern <a href="https://www.techtarget.com/searchitoperations/tip/Top-observability-tools">observability tools</a> provide comprehensive real-time visibility into application performance, operational health and resource utilization across hybrid environments.</li> <li><b>Invest in network mapping and topology tools. </b><a href="https://www.techtarget.com/searchnetworking/feature/network-mapping-tools-to-optimize-IT-infrastructure">Network mapping tools</a><b> </b>can automatically help detect hardware -- such as routers, switches and servers -- and create diagrams showing how they interact. Because the cloud is dynamic, continuous discovery is necessary to keep maps up to date as new deployments occur.</li> <li><b>Build a security-aware visibility architecture.</b> Both the networking and security teams should be using security frameworks to identify loopholes in security tooling and reduce the effects of limited network visibility.</li> </ol> <p><i>Wisdom Ekpotu is a DevOps engineer and technical writer focused on building infrastructure with cloud-native technologies.</i></p> </section> Digital transformation promises faster decision- making and greater UX. However, a lack of visibility into the network can hinder broad digital transformation goals. https://cdn.ttgtmedia.com/rms/onlineimages/disaster_recovery_a257795847.jpg https://www.techtarget.com/searchnetworking/tip/How-limited-network-visibility-impedes-digital-transformation Tue, 07 Jul 2026 15:00:00 GMT <p>AI continues to accelerate the shift from legacy, manual network management to self-optimized, self-healing environments. As enterprises adopt AI to improve speed, efficiency and resilience, CIOs and other IT leaders must shoulder new oversight responsibilities.</p> <p>AI-driven networking requires governance frameworks and automation guardrails to balance innovation with control and human accountability. This article examines how AI changes network operations, identifies common governance challenges, outlines risks and provides a roadmap for AI-driven network governance.</p> <section class="section main-article-chapter" data-menu-title="How AI is changing network operations"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How AI is changing network operations</h2> <p><a href="https://www.techtarget.com/searchcloudcomputing/tip/Ways-to-use-AI-for-cloud-infrastructure-management">AI transforms network operations</a> from reactive management to predictive and increasingly autonomous control. Practical uses include the following:</p> <ul class="default-list"> <li><b>AI-driven traffic management.</b> Improves bandwidth allocation and network performance.</li> <li><b>Machine learning.</b> Supports predictive maintenance and anomaly detection.</li> <li><b>AI automation.</b> Automates routine tasks, such as troubleshooting and configuration management. It also enables <a href="https://www.techtarget.com/searchNetworking/tip/AI-driven-self-healing-networks-bring-new-capabilities">self-healing networks</a> and real-time scaling.</li> </ul> <p>These benefits improve uptime, response times and scalability, which lets network teams focus on strategic initiatives rather than day-to-day maintenance.</p> </section> <section class="section main-article-chapter" data-menu-title="Governance challenges in AI-driven networks"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Governance challenges in AI-driven networks</h2> <p>As AI takes on a larger role in network operations, executive leadership must address the governance issues arising from increased automation. Without clear oversight, AI-driven environments can introduce uncertainty, reduce transparency and <a href="https://www.mergeworld.com/insights/ai-isnt-creating-gaps-its-revealing-them">create gaps</a> between technology decisions and broader business objectives.</p> <p>Specific challenges include the following:</p> <ul class="default-list"> <li><a href="https://www.techtarget.com/searchenterprisedesktop/feature/The-next-enterprise-AI-problem-is-visibility">Lack of visibility into AI</a> decision-making processes.</li> <li>Difficulty assigning accountability when automated systems make changes.</li> <li>Insufficient explainability for IT leaders to understand or justify AI-driven actions during outages, audits or compliance reviews.</li> <li>Risk of inconsistent policy enforcement across distributed network environments.</li> <li>Potential compliance or regulatory concerns.</li> <li>Disconnects between network, security and executive leadership teams.</li> </ul> <p>Lack of visibility and explainability erodes trust among executives and operations teams, while the complexity of autonomous systems creates risks.</p> </section> <section class="section main-article-chapter" data-menu-title="Risks of AI in networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Risks of AI in networking</h2> <p>While AI-driven networking can improve efficiency and responsiveness, it also increases the <a href="https://www.techtarget.com/searchenterpriseai/tip/AI-risk-management-A-strategic-guide-for-enterprise-leaders">risks that organizations must actively manage</a>. IT leaders face growing challenges around security, visibility and control over critical infrastructure.</p> <p>Specific risks include the following:</p> <ul class="default-list"> <li>Poorly governed AI systems can introduce security and availability vulnerabilities.</li> <li>Overreliance on automation reduces human understanding of network configuration and behavior.</li> <li>Hybrid and multi-cloud environments increase complexity and reduce centralized visibility.</li> <li>AI systems could make <a href="https://www.techtarget.com/searchcio/feature/AI-failure-examples-What-real-world-breakdowns-teach-CIOs">incorrect or unintended changes</a> that escalate outages or disrupt business operations.</li> </ul> <p>Without mature governance, organizations can lose control. The speed and scale of AI-driven decision-making can amplify problems without proper oversight. Governance balances agility and control, ensuring humans can override automation.</p> </section> <section class="section main-article-chapter" data-menu-title="How to establish governance for AI-driven networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to establish governance for AI-driven networking</h2> <p>Governance must be central to automation. CIOs and IT leaders need practical frameworks that balance innovation with accountability, transparency and control.</p> <p>Use the following governance roadmap to deploy AI-driven networking technologies while reducing risk, maintaining visibility and ensuring alignment with broader business objectives, security requirements and compliance obligations.</p> <ol class="default-list"> <li><b>Define ownership and accountability. </b>Establish clear governance roles across IT operations, cybersecurity and executive leadership. Assign responsibility for approving AI-driven actions, managing policies and handling incidents involving automated systems.</li> <li><b>Establish automation guardrails. </b>Define which network functions run autonomously, and which need human review or approval. Set escalation thresholds for high-impact changes, security events or unexpected network behavior to <a href="https://www.techtarget.com/searchnetworking/tip/Why-AI-in-network-operations-requires-human-judgment">ensure humans remain in the loop</a>.</li> <li><b>Improve visibility and monitoring.</b> Implement continuous monitoring, auditing and reporting. Keep detailed logs and performance metrics to help track automated decisions, support compliance audits and identify anomalies quickly.</li> <li><b>Prioritize explainability and transparency.</b> Require AI networking platforms to provide explainable decision-making capabilities. Transparency into how AI systems manage traffic, anomalies and remediation builds trust among network, operations, executives and auditors.</li> <li><b>Align governance with business objectives. </b>Regularly review AI policies to ensure they directly support broader enterprise priorities, including compliance, security, resilience and customer experience.</li> </ol> <p>Governance policies must evolve alongside AI capabilities and remain integrated with the overall business strategy rather than operate in isolation or without leadership.</p> </section> <section class="section main-article-chapter" data-menu-title="Governance is essential in networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Governance is essential in networking</h2> <p>AI-driven networking offers major advantages, but sustainable success requires governance that keeps automation transparent, accountable and aligned with business priorities.</p> <p>As AI assumes greater control of enterprise networks, CIOs and executive leaders who establish governance now will be better positioned to balance automation, resilience and accountability before complexity outpaces oversight.</p> <p><em>Damon Garn owns Cogspinner Coaction and provides freelance IT writing and editing services. He has written multiple CompTIA study guides, including the Linux+, Cloud Essentials+ and Server+ guides, and contributes extensively to TechTarget Editorial, The New Stack and CompTIA Blogs.</em></p> </section> AI transforms network operations through automation and predictive control, but requires strong governance frameworks to balance innovation with accountability and transparency. https://cdn.ttgtmedia.com/rms/onlineimages/collab_a275903017.jpg https://www.techtarget.com/searchnetworking/tip/How-CIOs-should-govern-AI-driven-network-operations Wed, 24 Jun 2026 17:00:00 GMT <p>A hyperscale data center is a facility that can excessively scale and quickly meet the needs of new, massive growth.</p> <p>These data centers can hold thousands of servers, are built to process more data than enterprise data centers and have a variety of features that make them unique. There are also challenges in designing and maintaining them compared to enterprise or smaller data centers, such as the need for advanced automation and high uptime requirements.</p> <section class="section main-article-chapter" data-menu-title="What is a hyperscale data center?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is a hyperscale data center?</h2> <p>By their nature, data centers are designed to operate at scale and often house dozens, if not hundreds, of physical servers and virtual machines (VMs). A <a href="https://www.techtarget.com/searchdatacenter/feature/Guide-to-understanding-the-various-types-of-data-centers">hyperscale data center</a> is essentially the next level up -- it can support thousands of servers and millions of VMs.</p> <p>The number of hyperscale data centers is rapidly growing. A&nbsp;<a href="https://www.grandviewresearch.com/industry-analysis/hyperscale-data-center-market" target="_blank" rel="noopener">market research report</a>&nbsp;from Grand View Research projects that the global hyperscale data center market will grow to $52.54 billion by 2030 at a compound annual growth rate of 13.6%.</p> <p><a href="https://www.techtarget.com/whatis/definition/hyperscale-computing">Hyperscale computing</a>&nbsp;takes a stripped-down approach to networking, combining compute, storage and virtualization layers into a single computing environment. The infrastructure networks servers horizontally to maximize the hardware, and a load balancer monitors the amount of data the facility needs to process.<br><br>Hyperscale data center architecture is designed to be exceptionally lean and agile. This enables the ability to share the processing load across the infrastructure and quickly add or remove servers or other resources as needed to meet capacity demands.</p> </section> <section class="section main-article-chapter" data-menu-title="Key features of a hyperscale data center"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Key features of a hyperscale data center</h2> <p>The approach to hyperscale architecture may vary from provider to provider, but most hyperscale data centers share a few common features, including the following:</p> <ul class="default-list"> <li><b>Large site locations.</b>&nbsp;Since hyperscale data centers can support thousands of servers, they tend to be much larger than the average data center. To be considered a hyperscale data center, the facility must occupy at least 10,000 square feet, with some of the truly massive facilities reaching hundreds of thousands or even millions of square feet.</li> <li><b>High-density server racks.</b>&nbsp;More square footage isn't the only way to fit more servers into a hyperscale facility. Most&nbsp;<a href="https://www.techtarget.com/searchdatacenter/feature/Learn-the-major-types-of-server-hardware-and-their-pros-and-cons">use specialized high-density server racks</a>. These server racks are wider and can accommodate more components, like power suppliers and hard drives, with enough space for engineers to swap them out and customize them as needed. Most hyperscalers have at least 5,000 servers.</li> <li><b>Strong energy resources.</b>&nbsp;With so much equipment to support, hyperscale facilities need a massive amount of power. Some facilities consume entire gigawatts of power, which is enough to power a small city. Hyperscalers&nbsp;<a href="https://www.techtarget.com/searchdatacenter/feature/Data-center-power-infrastructure-essentials-prevent-downtime">invest in the most advanced power</a>&nbsp;suppliers and HVAC systems for more efficient cooling. Hyperscale data centers are also often built in areas with cheap electricity, away from critical power grids.</li> <li><b>Advanced heat control</b>. High-density server racks come with their own set of cooling challenges. More servers, more components, and more complex computations to support AI workloads generate a lot of heat, and as such, many hyperscalers are equipped with state-of-the-art <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-heat-reuse-How-to-make-the-most-of-excess-heat">heat management systems</a>. These systems often use a combination of air circulation, liquid cooling and thermal optimization to maintain safe operating temperatures for data center equipment.</li> <li><b>Disaggregation and modularity.</b>&nbsp;The ability to customize and swap out components is key to the flexibility of hyperscale facilities and equipment. This requires a commitment to personalized configuration, which is expensive, but the result is highly improved modularity, making it easier to adjust infrastructure and equipment on short notice.</li> <li><b>Intelligent automation.</b>&nbsp;Given the sheer scale of these data centers, manual monitoring is not feasible. Many hyperscalers heavily rely on <a href="https://www.techtarget.com/searchdatacenter/tip/Balancing-automation-with-human-oversight-in-AI-data-centers">automation tools</a> to allocate assets, optimize workloads, monitor and repair systems, and more.</li> <li><b>AI workload support</b>. Hyperscalers are uniquely positioned to meet the demand for AI. <a href="https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-next-big-shifts-in-ai-workloads-and-hyperscaler-strategies" target="_blank" rel="noopener">Modular and prefabricated server builds</a> help hyperscalers handle AI workloads. These builds enable hyperscale facilities to scale up and out based on demand. In addition, more hyperscalers are consolidating AI workloads into <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-trends-to-watch">data center campuses</a> to more efficiently support the power and cooling needs of high-density server racks.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Benefits of a hyperscale data center"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Benefits of a hyperscale data center</h2> <p>These features translate to a range of practical benefits compared to traditional data centers:</p> <ul class="default-list"> <li><b>Flexibility and scalability.</b>&nbsp;The biggest advantage of hyperscalers is the ability to scale out both horizontally and vertically. This enables more balanced workloads and better resource provisioning. As demand increases or decreases, it's easy for hyperscale data centers to&nbsp;<a href="https://www.techtarget.com/searchdatacenter/tip/Best-practices-for-data-center-network-optimization">adjust and match that demand</a>, no matter the scale.</li> <li><b>Reduced downtime.</b>&nbsp;Since hyperscalers have a strong focus on automation and a large number of resources, downtime is reduced. The environments are so controlled that they can automatically recover from various issues. If there's a sudden spike in demand, the hyperscaler can meet those needs. Many redundancies are also in place to maintain uptime at all costs.</li> <li><b>Increased efficiency.</b>&nbsp;From operations to cooling, hyperscale data centers are designed to operate at peak efficiency. And&nbsp;<a href="https://www.techtarget.com/searchnetworking/answer/Whats-automations-role-in-network-provisioning-resources">thanks to increased automation</a>, they may not require as much staff to manage.</li> <li><b>Advanced technology.</b>&nbsp;As larger organizations evolve and iterate on hyperscale architecture, the technology and best practices they pioneer will hopefully trickle down to smaller data centers. This will, in turn, empower all data centers to use their physical space more effectively, provision resources more rapidly and reduce their power usage.</li> <li><b>Complex computational workloads</b>. As AI proliferates, so too does the need for complex computational workloads. To <a href="https://www.techtarget.com/searchenterpriseai/feature/How-to-build-a-machine-learning-model-in-7-steps">train an AI model</a>, specialized hardware architectures are needed to perform intense mathematical calculations millions of times. AI queries and inference workloads also require specialized network designs to support high interconnectivity and speed at scale. Hyperscale data centers can more easily meet these needs than traditional data centers, which aren't built to operate at that level.</li> <li><b>Pay-as-you-go pricing.</b>&nbsp;For hyperscale customers, the services can be more flexible. A hyperscale data center can sustain high service levels and scale as demand increases.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Challenges of hyperscale data centers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Challenges of hyperscale data centers</h2> <p>Many hyperscaler features also represent the biggest challenges, including space, power usage and advanced technologies.</p> <h3>Land and climate restrictions</h3> <p>Hyperscale data centers require a lot of land in areas with cheap, reliable electricity. A hyperscale data center should also not be anywhere with severe inclement weather&nbsp;<a href="https://www.techtarget.com/searchdisasterrecovery/tip/Underscore-preparation-in-your-natural-disaster-recovery-plan">that could bring the whole place down</a>&nbsp;-- not to mention the need for layered security to prevent cyber attacks and breaches that could do the same.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/pros_and_cons_of_hyperscale_data_centers-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/pros_and_cons_of_hyperscale_data_centers-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/pros_and_cons_of_hyperscale_data_centers-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/pros_and_cons_of_hyperscale_data_centers-f.png 1280w" alt="chart of benefits and challenges of hyperscale data centers" height="358" width="559"> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <h3>Difficulties in configuration</h3> <p>Customization is also a challenge. Hyperscale computing environments are still maturing. They require many aspects to be personalized and configured, which requires significant upfront expertise. Organizations that build hyperscale data centers also need to invest in extensive R&amp;D, in addition to advanced hardware and cabling. While they can remove the need for human expertise over the long term, modern automation tools can be expensive and difficult to implement.</p> <h3>Risk of overbuilding or underbuilding</h3> <p>It is possible to&nbsp;<a href="https://www.techtarget.com/searchdatacenter/How-to-design-and-build-a-data-center">overbuild or underbuild</a>&nbsp;a data center. In the former case, organizations could have idling resources, underused machines or even equipment that was new at the time but outdated by the time it is used. Underbuilding could result in overloaded machines and large-scale system failures.</p> <h3><a name="_heading=h.jdbekaxy9syi"></a>Increased energy and water stress</h3> <p>Traditional data centers have come under scrutiny for both <a href="https://www.techtarget.com/searchdatacenter/tip/How-much-energy-do-data-centers-consume">energy usage</a> and <a href="https://www.techtarget.com/searchdatacenter/tip/How-to-manage-data-center-water-usage-sustainably">water consumption</a>, and hyperscalers grapple with the same issues, even more so. While hardware advancements and software optimizations can help reduce energy use, and innovative cooling solutions can relieve water stress, hyperscale data centers still require immense amounts of power and water to function. Hyperscalers may need to build and invest in their own power grids and circular water systems to combat energy stress and water scarcity.</p> </section> <section class="section main-article-chapter" data-menu-title="Examples of existing hyperscale data centers"> <h2 class="section-title"><i class="icon" data-icon="1"></i><a name="_heading=h.7uqb4i22a9rh"></a>Examples of existing hyperscale data centers</h2> <p>According to <a href="https://www.srgresearch.com/articles/hyperscale-data-center-count-hits-1136-average-size-increases-us-accounts-for-54-of-total-capacity" target="_blank" rel="noopener">Synergy Research Group</a>, more than half of the world's hyperscale data center capacity is in the U.S., with Europe and China following, each with about a third of the remaining worldwide capacity. The largest facilities belong to cloud providers, including Microsoft, Amazon and Google. Smaller hyperscale operators include Meta, Alibaba, Tencent and Apple, among others.</p> <p>There were&nbsp;at least 1,136 hyperscale data centers at the end of 2024, with Synergy Research Group forecasting an additional 130 to 140 new hyperscale data centers to come online annually over the next few years.</p> <p>The Chinese market is rapidly growing and currently has some of the largest hyperscale data centers. For example, the Inner Mongolia Information Hub facility in Hohhot, China, is 10.7 million square feet and is one of the largest data centers in the world. There's also the Range International Information Hub in Langfang, China, which is 6.6 million square feet.</p> <p>In the wake of the data center and AI boom, a new public-private AI infrastructure initiative, <a href="https://openai.com/index/announcing-the-stargate-project/" target="_blank" rel="noopener">The Stargate Project</a>, was announced in the U.S. The goal is to meet demand by building a nationwide network of AI hyperscalers and data center campuses, the first of which is being built in Abilene, Texas. OpenAI will oversee operations of this facility, but the initiative has received funding from several organizations, including the U.S. government, SoftBank, Oracle and MGX. Arm, Microsoft and Nvidia have joined the project as key technology partners.</p> </section> <section class="section main-article-chapter" data-menu-title="Keep up with developments to prepare for hyperscaling"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Keep up with developments to prepare for hyperscaling</h2> <p>There are considerable challenges to supporting hyperscale computing in data centers, which means only the biggest players in the field will likely be able to afford, build and maintain these facilities. Plus, certain regions and states are <a href="https://www.ncsl.org/fiscal/which-states-are-banning-data-centers" target="_blank" rel="noopener">introducing legislation</a> to ban or limit the construction of new hyperscalers to protect local communities from the potential impacts of data centers and improve power grid resiliency.</p> <p>However, data center administrators should stay up to date on the latest advancements and trends in hyperscale computing as the technology and <a href="https://www.techtarget.com/searchstorage/tip/Apply-hyperscale-data-center-design-principles-to-IT-storage">best practices</a> trickle down. There may also be opportunities for operators to consolidate data center sites with hyperscale campuses, which consist of multi-facility communities that can replicate data and failover processes to provide more flexible and reliable services.</p> <p>As the <a href="https://www.techtarget.com/searchenterpriseai/feature/Is-the-AI-bubble-about-to-burst-or-is-it-recalibrating">AI boom</a> continues and data demand grows, the world's data center infrastructure must grow in kind. Hyperscale data centers and campuses that can support massive power output on their own electrical grids and land will become necessary. Hyperscalers will also require more advanced cooling techniques and technologies, modular hardware and scalable architecture, and sustainable infrastructure to meet demand over the short and long term.</p> <p>Rapid growth is accelerating the future of hyperscale, and by keeping up with its rapid evolution, all data centers can benefit from the pioneers pushing the field forward.</p> <p><b>Editor's note: </b>This article was updated in 2026 to add up-to-date information about hyperscalers and to improve the reader's experience.</p> <p><em>Jacob Roundy is a freelance writer and editor with more than a decade of experience with specializing in a variety of technology topics, such as data centers, business intelligence, AI/ML, climate change and sustainability. His writing focuses on demystifying tech, tracking trends in the industry, and providing practical guidance to IT leaders and administrators.</em></p> </section> Hyperscale data centers can hold thousands of servers and process much more data than an enterprise facility. However, they can be difficult to build and maintain. https://cdn.ttgtmedia.com/rms/onlineimages/storage_g1180684429.jpg https://www.techtarget.com/searchdatacenter/tip/A-primer-on-hyperscale-data-centers Fri, 05 Jun 2026 16:15:00 GMT <p>AI is continuously changing, evolving and upgrading. While these advancements benefit organizations, they put a strain on the networking infrastructure.</p> <p>During a keynote at Cisco Live 2026, Denise Lee, vice president of engineering sustainability at Cisco, evaluated this challenge and presented Cisco's response: Energy Networking Systems. This article discusses the characteristics of this three-pillar <a href="https://www.techtarget.com/searchnetworking/tip/How-network-efficiency-advances-ESG-goals">networking approach to conserving energy</a> and costs in the AI era, and explores how Workday tested the architecture in a real enterprise environment.</p> <section class="section main-article-chapter" data-menu-title="Rising energy demands in the AI era"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Rising energy demands in the AI era</h2> <p>According to Lee, components such as switchboards, power, cooling and electrical infrastructure, also known as <i>gray space</i>, are becoming increasingly important in data centers in the AI era. For example, data center capacity is projected to increase to 200 gigawatts by 2030, consuming 300% more energy by 2050, according to research from National Electrical Manufacturers Association’s 2025 <a href="https://www.makeitelectric.org/about/our-story/year-in-review/">Year in Review</a>.</p> <p>"Artificial intelligence is rewriting what's possible, and it's clear our energy systems must evolve to keep pace," she said. "Every model, every breakthrough demands more power than the last."</p> <p>The increased power requirements have made it more difficult to build a data center, Lee said. In the past, when operators built data centers, the assumption was that sufficient power was available for deployment. Now, in the age of AI, data center operators must consider physical characteristics -- such as grid capacity, water supply and infrastructure build-out time -- as these constraints can <a href="https://www.techtarget.com/searchdatamanagement/feature/Power-constrained-data-architecture-curbing-AI-ambitions">hinder the ability to build a plant</a>.</p> </section> <section class="section main-article-chapter" data-menu-title="Cisco's three-pillar framework"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Cisco's three-pillar framework</h2> <p>Energy Networking Systems is Cisco's answer to this power problem. Rather than an architecture platform, Energy Networking Systems is an umbrella term that encompasses the following three main components for managing power-hungry AI-native infrastructure:</p> <ol class="default-list"> <li>Energy visibility.</li> <li>Power distribution.</li> <li>Thermal management.</li> </ol> <h3>Energy visibility</h3> <p>The first layer of Energy Networking Systems is energy visibility, provided through a dashboard in Cisco Cloud Control called Energy Management. It provides <a href="https://www.techtarget.com/searchstorage/tip/How-telemetry-can-transform-storage-strategies">data center operators with the telemetry data</a> necessary for the following:</p> <ul class="default-list"> <li>Visibility and reporting.</li> <li>Insights and analytics.</li> <li>Control and automation.</li> </ul> <p>The tools within Energy Management enable organizations to understand their energy consumption, mix and costs, while recognizing the carbon intensity and greenhouse gas emissions of their projects.</p> <h3>Power distribution</h3> <p>Cisco is supporting power distribution with modern technology called Fault Managed Power (FMP). This technology, developed by engineers worldwide, including those at Cisco, delivers more electricity over longer distances. Unlike traditional electricity delivery systems that require the data center to convert energy before it reaches the devices, FMP uses transmitters, cables and receivers to send electricity to <a href="https://www.techtarget.com/searchdatacenter/Server-hardware-guide-to-architecture-products-and-management">devices in a data center rack</a>.</p> <p>"It is safe to touch, faster to install, requires less material and the limits on wattage are sort of left up to the imagination because it goes up to 450-volt DC," Lee said.</p> <p>Cisco's focus on power distribution suggests that it is critical to AI data center networking. This improved form of power distribution can not only power AI data centers, but also entire buildings -- including smart buildings -- edge-computing equipment, battery energy storage systems and more.</p> <h3>Thermal management</h3> <p>Cooling is one of the most important parts of this umbrella, because so many resources go toward it. In traditional air-cooled data centers, more than 40% of electricity goes toward cooling. Lee described this as "non-useful power," meaning that a large amount of energy is being directed toward a less significant resource.</p> <p>Many newer cooling methods are available, such as rear-door heat exchangers or direct-to-chip cooling. Lee noted some of Cisco's cooling approaches, including its network switches with direct <a href="https://www.techtarget.com/searchdatacenter/tip/Explore-liquid-cooling-for-data-centers">liquid cooling</a> -- such as the Cisco Silicon One G300 102.4T switch -- and immersion cooling.</p> <p>"When you look at all these different emerging architectures for power and cooling, specifically with cooling, it's a double knock-on effect because you no longer have to spend as much power [on] cooling," Lee said. "And that cooling may not take as much power to get the output you need."</p> <p>With these different options, Lee said, organizations now have multiple choices for cooling. One data center can have different cooling options. This is especially useful in situations where organizations must retrofit their existing data centers.</p> </section> <section class="section main-article-chapter" data-menu-title="Workday's phased approach to FMP"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Workday's phased approach to FMP</h2> <p>A power distribution technology like FMP was especially useful to Workday, according to Ben Paterson, a technology and smart building leader at the company. Workday faced visibility concerns, unable to recognize how much power the organization was consuming.</p> <p>"A facility needs to be a living and breathing entity," Paterson said. "Everything should provide data [so that] we can track how people are utilizing the space, track flow, or heating and cooling."</p> <p>Paterson and his team decided to deploy FMP in a <a href="https://www.techtarget.com/searcherp/feature/Enterprise-platforms-are-evolving-piece-by-piece">phased approach</a>. First, they created a proof-of-concept with Cisco and its partners to see how FMP could benefit them. In the first phase, they installed a temporary setup with FMP in a space within a facility. During this test, FMP powered everything in the room, such as the lights, desk power and monitoring equipment. In a typical electrical distribution system, this would have taken Workday weeks, Paterson said. With FMP, it only took an hour or two.</p> <p>Now in phase two of the deployment, Workday is expanding FMP across its entire office environment and testing how to improve it within the organization. Paterson added that if he were to build a new data center from the ground up, he would consider how much power his networking equipment and servers use. When these consume less power, the organization experiences less power loss.</p> </section> <section class="section main-article-chapter" data-menu-title="The future of data center efficiency"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The future of data center efficiency</h2> <p>Only a minority of power in data centers goes toward useful compute, according to Cisco research. In addition to the 40% of non-useful power that goes toward cooling in data centers, another 25% is lost in power distribution, bringing the total to 65%. The annual energy cost of running one of these data centers averages around $6 million.</p> <p>However, according to the same report, <a href="https://www.techtarget.com/searchdatacenter/feature/Liquid-cooling-vs-air-cooling-in-the-data-center">switching from air cooling to liquid cooling</a> can reduce non-useful power by more than 20%, bringing the total to 43%. Organizations can also see a half reduction in costs, going from $6 million to $3 million.</p> <p>Organizations can see even more savings when combining liquid cooling with FMP. Non-useful energy reduces to less than a quarter at 22%, and the costs of running a data center drop to $2 million. This can result in 76% in energy cost savings and enable more than double the compute with the same energy.</p> <p>With that in mind, however, every setup will be different. Similar to Workday's phased deployment, organizations will need to test Energy Networking Systems technologies to find out how they fit into their existing data center infrastructure.</p> <p>"The actual benefit of efficiency and how you design, wire [and] architect it is up to you because it varies," Lee said. "Every environment is slightly different. We're encouraging people to get into your environments, get a proof-of-concept going [and] get a proof-of-value going."</p> <p><em>Deanna Darah is site editor for TechTarget's Networking site. She began editing and writing at TechTarget after graduating from the University of Massachusetts Lowell in 2021.</em></p> </section> AI's growth is pushing data centers to their energy limits. Cisco's Energy Networking Systems is a three-pillar approach to manage power, cooling and costs in AI infrastructure. https://cdn.ttgtmedia.com/rms/onlineimages/ai_a252657224.jpg https://www.techtarget.com/searchnetworking/feature/Energy-driven-networking-trends-emerging-in-the-AI-era Fri, 05 Jun 2026 15:30:00 GMT <p>Cisco put agentic AI in the spotlight at its 2026 conference, held May 31-June 4 in Las Vegas.</p> <p>Keynotes from Cisco leadership at the event focused on how the tech giant is helping enterprises create, secure and expand their agentic AI infrastructure. Jeetu Patel, Cisco president and chief product officer, also highlighted the importance of enterprises being able to tackle what he called the "AI trust deficit" as use of AI agents expands.&nbsp;</p> <p>At keynote "Deep Dive" sessions, conference-goers had the opportunity to learn more about Cisco's agentic initiatives in areas such as AgenticOps for autonomous network management.</p> <p>Informa TechTarget's editorial team was on site at the event, providing coverage of Cisco's major announcements. Use this guide to catch up on all the news, analysis and insights from the show.</p> Informa TechTarget's updating guide to Cisco Live 2026 covers the latest trends, emerging technologies and product announcements from the tech giant's annual conference. https://www.techtarget.com/searchnetworking/conference/Cisco-Live-conference-coverage-news-and-analysis Tue, 02 Jun 2026 00:00:00 GMT <p>Many IT organizations are turning to AI as they plan to replace the tools they now use to monitor and troubleshoot their networks.</p> <p>Enterprise Management Associates (EMA) recently surveyed 352 network management professionals for the 2026 edition of its biennial research report, <a href="https://www.enterprisemanagement.com/product/network-management-megatrends-2026-automation-hybrid-and-multi-cloud-networks-and-ai-transformation/">Network Management Megatrends</a>. The research found that 33% of IT organizations are very likely to replace their network observability tools over the next two years. In 2024, only 26% felt that strongly about tool replacement. Another 40% in this year's report said they were somewhat likely to replace a tool.</p> <section class="section main-article-chapter" data-menu-title="NetOps pros consider AI as they seek better tools"> <h2 class="section-title"><i class="icon" data-icon="1"></i>NetOps pros consider AI as they seek better tools</h2> <p>On the surface, the reason for the increase in tool churn is obvious. Only 32% of networking pros told EMA they were completely satisfied with their current tools. The average research participant said better tools would proactively prevent 53% of network faults and performance issues they now face.</p> <p>EMA asked respondents what would most motivate them to replace a network observability tool. Fifty-four percent pointed to <a href="https://www.techtarget.com/searchnetworking/tip/AI-driven-network-management-tasks">AI-driven insights</a> and automation. Notably, 55% of respondents also told EMA that AI features are a requirement when they evaluate network management tools in general.</p> <p>"We definitely want something agentic that is actively working on problems," a network tools lead with a multinational bank recently told EMA. "I don't see us changing vendors in the next fiscal year, but it's something we would do in the future."</p> <p><a href="https://www.enterprisemanagement.com/product/ai-driven-netops-how-enterprises-are-embracing-intelligent-network-management-solutions-2/">Previous EMA research</a> found that network pros would most like to apply AI to the following aspects of day-to-day network operations:</p> <ul class="default-list"> <li><b>Problem detection</b>. Fifty-eight percent of network teams want AI to help them proactively detect network trouble before it impacts the business. This enables them to reduce downtime and service disruptions, a key measure of NetOps success.</li> <li><b>Alert management</b>. Fifty-three percent of network teams believe AI can streamline and improve the process of configuring alert policies and thresholds, while also <a href="https://www.techtarget.com/searchsecurity/feature/How-to-improve-the-SOC-analyst-experience-and-why-it-matters">reducing noise</a> and enriching alerts with actionable information.</li> <li><b>Problem resolution</b>. Fifty-one percent of network teams believe newer agentic capabilities from tool vendors can troubleshoot problems and identify fixes. AI can present this remediation as a suggestion or act automatically without human involvement.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Other drivers of tool replacement"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Other drivers of tool replacement</h2> <p>AI isn't the only reason network teams are looking for new tools. Approximately 54% of survey respondents said better end-to-end visibility across networks, applications, clouds and UX would motivate them to replace network observability software. In fact, NetOps teams that collaborate more frequently with cloud and DevOps teams were more likely to want a new tool.</p> <p>Some 47% of respondents wanted their observability software to offer stronger support of modern architectures such as hybrid and multi-cloud networks and <a href="https://www.techtarget.com/searchnetworking/definition/Secure-Access-Service-Edge-SASE">secure access service edge</a>. EMA also found that respondents who intended to unify networking across their on-premises and cloud networks were more likely to replace their tools.</p> <p>Finally, EMA research found that a focus on automation motivates tool turnover. For example, 26% of network managers said that automation of <a target="_blank" href="https://octopus.com/blog/difference-between-day-0-1-2-operations" rel="noopener">Day 2</a> network operations -- such as event management, troubleshooting and optimization -- is a high priority for their organizations. These respondents were especially likely to replace a network observability tool. Many of them are discovering that the automation capabilities of incumbent vendors are inadequate.</p> <p>That said, switching monitoring tools isn't necessarily easy. Even though network observability vendors tend to push subscriptions over perpetual licenses today, many enterprises commit to three- or five-year subscriptions, <a href="https://www.techtarget.com/searchdatacenter/definition/vendor-lock-in">which locks them in</a>. In addition, many observability tools are highly customized and deeply integrated into other IT systems and operational procedures, making them sticky. This is especially the case in larger companies.</p> <p>Regardless, the will to change is there, and that drive will ultimately prevail. If an incumbent vendor isn't innovating in this new era of agentic AI, network operations teams will look elsewhere for one that does.</p> <p><em>Shamus McGillicuddy is vice president of research for the network management practice at Enterprise Management Associates (EMA). He has more than 20 years of experience in the IT industry and has written extensively about the network infrastructure market. Prior to joining EMA, McGillicuddy was the news director for TechTarget's networking site.</em></p> </section> Network managers want more from their observability tools and are looking at agentic AI to give them the flexibility and oversight they need. https://cdn.ttgtmedia.com/rms/onlineimages/toolGearArrow_g473747386.jpg https://www.techtarget.com/searchnetworking/post/How-AI-innovation-is-driving-network-observability-tool-churn Thu, 21 May 2026 13:15:00 GMT <p>No matter the level, most network professionals feel the same way about AI. From networking architects to day-to-day engineers, network professionals experience the same struggles and successes with AI in network operations.</p> <p>As the field of AI evolves rapidly, it's important to understand how network professionals are integrating the technology into operations. Mark Leary, research director of network analytics and automation at IDC, shared some of these insights from an April 2026 IDC report, based on responses from 516 respondents, during a session at ONUG's Spring 2026 AI Networking Summit.</p> <p>This article outlines 10 insights about how network professionals implement AI in network operations in 2026.</p> <section class="section main-article-chapter" data-menu-title="1. Network barriers that impede AI success"> <h2 class="section-title"><i class="icon" data-icon="1"></i>1. Network barriers that impede AI success</h2> <p>In a previous IDC survey about AI project hindrances conducted around two years ago, Leary said that networking ranked at the very bottom of the list of <a href="https://www.techtarget.com/searchenterpriseai/feature/AI-deployments-gone-wrong-The-fallout-and-lessons-learned">impediments to AI success</a>. At that time, AI projects were less common because the technology had not yet developed to a sufficient level.</p> <p>"Now, as we've seen people moving from pilots to production, all of a sudden the network matters," Leary said.</p> <p>Networking moved up the list in a survey conducted around 18 months ago. In IDC's most recent survey, it ranked fourth. As networking rises on the list, he said, staffing has dropped from the top three to the fifth-place spot. Organizations are less concerned with staffing issues as they continue to hire professionals and AI tools become easier to use.</p> <p>Barriers to AI success specific to the network include the following:</p> <ul class="default-list"> <li>Network security concerns.</li> <li>Network automation challenges.</li> <li>Staffing shortages.</li> <li>Cloud connectivity limitations.</li> <li>Limited network visibility and control.</li> <li>Unaffordable network upgrades.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="2. Platform vs. best-in-class"> <h2 class="section-title"><i class="icon" data-icon="1"></i>2. Platform vs. best-in-class</h2> <p>When IDC conducted the same study two years ago, two-thirds of respondents said they preferred platform approaches over best-in-class, Leary said. Top reasons for this preference include the following:</p> <ul class="default-list"> <li>Strengthened security.</li> <li>Reduced complexity.</li> <li>Lowered costs.</li> <li>Faster deployment.</li> <li>Easier integration.</li> </ul> <p>However, interest in a platform-based approach has declined since that survey. Now that network professionals are more experienced and comfortable with AI, only 45% of respondents said they prefer integrated technologies from a single provider. A best-in-breed approach is now leading, with 55% preferring a mix of technologies from various providers.</p> </section> <section class="section main-article-chapter" data-menu-title="3. Machine vs. human execution"> <h2 class="section-title"><i class="icon" data-icon="1"></i>3. Machine vs. human execution</h2> <p>Most network professionals prefer an AI-powered approach to network management, but a <a href="https://www.techtarget.com/searchhrsoftware/news/366593543/Humans-in-the-loop-wont-prevent-AI-disasters-experts-say">human-in-the-loop approach</a> is gaining traction. An approximate 87% of respondents said they preferred AI-powered network management tools for remediation and optimization:</p> <ul class="default-list"> <li>46% of respondents said they want tools that determine and execute actions.</li> <li>41% of respondents said they want tools that guide actions, but don't execute them automatically.</li> </ul> <p>Respondents reported using AI-powered automation for tasks such as threat response, configuration management and validation, and network problem diagnosis.</p> </section> <section class="section main-article-chapter" data-menu-title="4. AI network management functions"> <h2 class="section-title"><i class="icon" data-icon="1"></i>4. AI network management functions</h2> <p>Respondents reported <a href="https://www.techtarget.com/searchnetworking/tip/AI-driven-network-management-tasks">using AI for several network management tasks</a>, including the following:</p> <ul class="default-list"> <li>Network data collection.</li> <li>Threat detection and response.</li> <li>Agentic AI for network engineering and operations.</li> <li>Network automation.</li> <li>Network traffic prioritization and optimization.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="5. AI for task automation by location"> <h2 class="section-title"><i class="icon" data-icon="1"></i>5. AI for task automation by location</h2> <p>Most network professionals support using AI to automate network engineering and operational tasks. However, the percentage of tasks automated by AI is lower than expected.</p> <p>In 2024, IDC found that the percentage of network tasks automated by network professionals <a target="_blank" href="https://medium.com/@sk9128/university-networking-design-59208a96bf15" rel="noopener">managing campus and branch networks</a>, as well as data center or cloud networks, averaged around 20%. However, this gap began to widen in 2025. While 31% of tasks were automated in campus and branch networks, only 23% of tasks were automated in data center and cloud networks.</p> <p>An approximate 54% of respondents in campus or branch networks said they plan to automate tasks in 2027. By the same timeframe, only 29% on the data center and cloud side said they plan to automate tasks.</p> <p>"[It] should start looking more like our expectations relative to campus and edge," Leary said. "Beware of stagnating in the data center and pushing forward on the campus and branch side of the fence."</p> </section> <section class="section main-article-chapter" data-menu-title="6. Networking domains for AI innovation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>6. Networking domains for AI innovation</h2> <p>Networking respondents reported several areas where they believe AI can improve the most. The top areas include the following:</p> <ul class="default-list"> <li>Cloud connectivity.</li> <li>Network security.</li> <li>Data center networking.</li> <li>Wireless WAN.</li> <li>IoT.</li> <li>Wireless LAN.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="7. Current and future AI networking priorities"> <h2 class="section-title"><i class="icon" data-icon="1"></i>7. Current and future AI networking priorities</h2> <p>Most networking professionals currently use AI for tool integration, data security, and data collection and analysis. In the future, however, network professionals plan to use AI for the following capabilities:</p> <ul class="default-list"> <li>Improve AI security.</li> <li>Increase the number of autonomous capabilities of an AI system.</li> <li>Reduce cost and complexity.</li> <li>Accelerate innovation.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="8. Agentic AI priorities"> <h2 class="section-title"><i class="icon" data-icon="1"></i>8. Agentic AI priorities</h2> <p>Respondents reported using <a href="https://www.techtarget.com/searchenterpriseai/feature/Real-world-agentic-AI-examples-and-use-cases">agentic AI to automate several tasks</a>, including the following:</p> <ul class="default-list"> <li>Network optimization.</li> <li>Security enforcement.</li> <li>Network data visualization.</li> <li>Troubleshooting.</li> <li>Configuration management.</li> <li>Compliance monitoring and management.</li> <li>Hardware and software deployment.</li> <li>Network design.</li> </ul> <p>However, despite interest in agentic AI for networking, many network professionals remain apprehensive about the <a href="https://www.techtarget.com/searchenterpriseai/feature/Security-risks-in-agentic-AI-systems-and-how-to-evaluate-threats">potential problems it could create</a>.</p> <p>"We've been very comfortable with the agent approach in networking forever," Leary said. "The difference from long ago is [that] agents were collectors, and they were passive. Today's agents, even in the network, are very active."</p> <p>Top concerns with agentic AI include the following:</p> <ul class="default-list"> <li>Security.</li> <li>AI agent deployment and management complexities.</li> <li>Integration challenges.</li> <li>Costs.</li> <li>Network data constraints.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="9. AI networking suppliers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>9. AI networking suppliers</h2> <p>Because AI in networking is so complex and critical for organizations, network professionals have certain expectations for supplier companies that partner with their teams. Examples of expectations include the following:</p> <ul class="default-list"> <li>Incorporate agentic AI into data systems.</li> <li>Prioritize security across AI deployments.</li> <li>Apply AI across the networking stack.</li> <li>Provide an ecosystem of AI technologies and partnerships.</li> <li>Offer AI training and consulting.</li> <li>Use AI within their own business and IT platforms.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="10. AI in networking benefits"> <h2 class="section-title"><i class="icon" data-icon="1"></i>10. AI in networking benefits</h2> <p>Despite AI being meant to support staff productivity, development and value, staff is often considered one of the biggest roadblocks to AI adoption, Leary said. Staff often push back on the use of AI within their systems, believing it will save their jobs. However, embracing AI could <a href="https://www.techtarget.com/searchapparchitecture/tip/How-AI-can-support-developer-productivity">boost performance and productivity</a>.</p> <p>When network professionals embrace AI, it could lead to the following benefits:</p> <ul class="default-list"> <li>Increased productivity and teamwork.</li> <li>Strengthened security posture.</li> <li>Faster problem detection, diagnosis and direction.</li> <li>Increased business agility.</li> <li>Improved digital experience for employees and customers.</li> <li>Shift from reactive network management to proactive network management.</li> </ul> <p>"It is undoubtedly a career-changing moment for the network engineer and network operator to embrace AI as opposed to repel against it," Leary said. "This defines success over the next 10 years for any network engineer or operator."</p> <p><i>Deanna Darah is site editor for Informa TechTarget's SearchNetworking site.</i>&nbsp;</p> </section> A recent survey revealed 10 key insights on AI adoption in networking. Research shows network professionals currently use AI for automation, threat detection and optimization. https://cdn.ttgtmedia.com/rms/onlineimages/iot_g1226985345.jpg https://www.techtarget.com/searchnetworking/feature/insights-on-AI-adoption-in-network-operations Mon, 18 May 2026 16:00:00 GMT <p>Many network professionals are skeptical about AI. However, AI is less of an antagonist and more of an assistant to network professionals, and it works far better when users work alongside AI rather than against it.</p> <p>During this week's Spring 2026 AI Networking Summit, presented by ONUG, IBM outbound product management leader Jason Lovelace presented on the importance of this partnership. AI is being used across every sector of work, but how network professionals engage with the technology differs from how other professionals use it, Lovelace said. Even if AI is accurate 80% of the time, that's not sufficient for networking.</p> <p>"To implement AI in the network, you need to think about maintaining or improving on three nines," Lovelace said. "The goal here is the illumination of human judgment through AI as a partner."</p> <p>To achieve this high availability, organizations shouldn’t give AI free rein in the network. Rather, Lovelace suggested network professionals maintain their judgment and <a href="https://www.techtarget.com/searchcio/tip/AI-augmented-teams-Training-for-human-machine-collaboration">work with AI as a partner</a>. He outlined this process through IBM's four-step framework:</p> <ol class="default-list"> <li>See: Gain temporal insights.</li> <li>Use: Apply reasonings and use tools.</li> <li>Prove: Evaluate suggestions and apply guardrails.</li> <li>Act: Execute the changes.</li> </ol> <p>This article outlines how network professionals can follow IBM's model to improve their AI use in networking.</p> <section class="section main-article-chapter" data-menu-title="1. See"> <h2 class="section-title"><i class="icon" data-icon="1"></i>1. See</h2> <p>This first step of the model concerns how AI observes the network and gathers information from the environment. According to Lovelace, an <a href="https://www.techtarget.com/searchenterpriseai/tip/How-to-build-your-first-agentic-AI-system">agentic AI system</a> needs two critical pieces to collect useful information about the network:</p> <ol class="default-list"> <li>Telemetry.</li> <li>Context.</li> </ol> <h3>Telemetry</h3> <p>Real-time telemetry data is essential for the AI tool to understand what happens in the network as soon as it occurs. Delays of up to a few minutes can become problematic because the system will no longer evaluate the network in real time.</p> <p>Lovelace suggested using <a target="_blank" href="https://medium.com/@learn-simplified/time-series-analysis-what-why-and-how-d01a208d2073" rel="noopener">time-series</a> monitoring -- an approach that collects metrics over time and uses past behavior to understand how a system will perform -- to evaluate telemetry data. Time-series tools can also spot issues within the network that occur before it passes the threshold before an alarm.</p> <p>According to Lovelace, combining agentic AI with time-series monitoring enables the AI to have a comprehensive understanding of the network, the next most essential part of analyzing data.</p> <h3>Context</h3> <p>When an agentic AI model has deep context about the network, teams can use it to analyze anomalies and other issues that occur in the network more thoroughly, Lovelace said.</p> <p>For example, network teams can use this metric data to determine whether an issue is isolated or points to a broader issue, such as distinguishing whether an alert is the issue itself or a symptom of the true problem.</p> <p>"We need both of these things working together to give engineers a clear view of what's happening," he said.</p> <p>However, Lovelace cautioned against an AI agent having too much context, as it could perform less accurately and <a href="https://www.techtarget.com/searchenterpriseai/feature/Security-risks-in-agentic-AI-systems-and-how-to-evaluate-threats">lead to more problems</a> than fixes.</p> </section> <section class="section main-article-chapter" data-menu-title="2. Use"> <h2 class="section-title"><i class="icon" data-icon="1"></i>2. Use</h2> <p>Network professionals next apply AI tools to the network. According to Lovelace, one of the most important things to recognize about this step is that the models are less important than the tools themselves.</p> <p>As models continue to evolve, the type of model used is less important than the ability to create and secure useful tools for network engineers, he said. <a href="https://www.techtarget.com/searchenterpriseai/tip/Compare-top-AI-coding-tools">AI tools that generate code</a> can especially help teams build as many tools as needed, giving network engineers a vast array of tools to support operations, he added.</p> <p>"The richness of the tools is critical for thinking about what your network engineers do on a daily basis," Lovelace said.</p> </section> <section class="section main-article-chapter" data-menu-title="3. Prove"> <h2 class="section-title"><i class="icon" data-icon="1"></i>3. Prove</h2> <p>This step is less about proving that AI works and more about <a href="https://www.techtarget.com/searcherp/feature/4-structural-foundations-of-reliable-enterprise-AI">demonstrating AI reliability</a>. Unlike other issues that AI tools might help remediate, network problems are more likely to require more prompts in an AI tool. The more requests submitted to an AI tool, the more complexity it has and the lower its accuracy.</p> <p>"If your context window is filling up, the compounding effect of a hallucination at step four or five then makes steps 17, 18 and 19 not reliable," Lovelace said.</p> <p>Because of the complexity of network operations, engineers must prevent AI tools from relaying inaccurate information down the chain of processes. Additionally, network teams must implement guardrails to prevent compounding errors and <a href="https://www.techtarget.com/searchnetworking/tip/How-to-ensure-network-performance-and-reliability">ensure the network's overall reliability</a>.</p> </section> <section class="section main-article-chapter" data-menu-title="4. Act"> <h2 class="section-title"><i class="icon" data-icon="1"></i>4. Act</h2> <p>The final step of the model defines how network engineers put AI into practice. Rather than offering a blanket approach to AI network operations, Lovelace said the ways network engineers use AI depend on their level of seniority.</p> <p>Junior network engineers would benefit from using AI to guide their decision-making. AI can suggest steps, and network engineers would follow and implement them while learning the explanations. A senior network engineer, on the other hand, should consult AI for recommendations and then decide which, if any, of its suggestions to implement.</p> <p>"They want to be able to have the ability to push back on the model and have the model reconsider its actions," Lovelace said.</p> </section> <section class="section main-article-chapter" data-menu-title="AI framework builds trust across engineering teams"> <h2 class="section-title"><i class="icon" data-icon="1"></i>AI framework builds trust across engineering teams</h2> <p>An AI framework could provide a means for AI use that network engineers and leaders are comfortable with. Additionally, it could help engineers improve their skills both directly within network operations and validate their judgment.</p> <p>For junior network engineers, an AI framework could help them understand the work they're doing in a network when analyzing, validating and using network data. This understanding could help a network engineer advance in network operations, from entry-level to mid-level to senior-level. For senior network engineers, an AI framework can help them build trust in AI models and enhance their&nbsp;judgment and understanding of network behavior.</p> <p>Finally, an AI usage framework can also ensure that leaders know their engineers are using AI safely, in line with <a href="https://www.techtarget.com/searchenterpriseai/tip/Agentic-AI-governance-strategies-A-complete-guide">governance and reliability requirements</a>.</p> <p>"If the AI technologies change, the framework allows organizations to reap the benefits of agentic NetOps and consume those changes while the standards of good governance and clean execution don't change," Lovelace said.</p> <p><i>Deanna Darah is site editor for Informa TechTarget's SearchNetworking site.</i></p> </section> Many network pros are skeptical about AI, but it's most effective as an assistant to human judgment. A framework helps engineers improve skills and build confidence. https://cdn.ttgtmedia.com/rms/onlineimages/ai_a252657224.jpg https://www.techtarget.com/searchnetworking/tip/Why-AI-in-network-operations-requires-human-judgment Mon, 18 May 2026 16:00:00 GMT <p>Despite concerns, AI is heavily embedded in network operations. However, this doesn't mean network professionals are losing control over their networks.</p> <p>Autonomous networking doesn't mean relinquishing full control of the network. Instead, it refers to the automated interpretation of network state <a href="https://www.techtarget.com/searchnetworking/tip/How-intelligent-network-automation-transforms-enterprise-operations">without human intervention</a>. Network engineers aren't being replaced -- they're <a href="https://www.techtarget.com/searchnetworking/feature/Networking-pros-face-strong-job-market-greater-demands">taking on a new role</a>.</p> <p>Modern AIOps systems are intelligent and continuously process telemetry data, logs, flows, metrics and events to detect patterns and surface actionable insights. This article discusses which processes AIOps systems can effectively automate in networking and how that <a href="https://www.techtarget.com/searchnetworking/post/How-network-engineers-can-prepare-for-the-future-with-AI">affects network management</a>.</p> <section class="section main-article-chapter" data-menu-title="The evolution of AI in network management"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The evolution of AI in network management</h2> <p>The evolution of network management can be outlined in the following three stages:</p> <p>1.<b> Manual operations. </b><a href="https://www.techtarget.com/searchnetworking/tip/Network-tasks-administrators-can-do-quicker-from-the-CLI">CLI-based troubleshooting</a>, reactive responses and error-prone processes. <b><br></b>2. <b>Automation. </b>Error-prone CLI troubleshooting became possible through automation. Network teams went from implementing manual processes to running scripted workflows and policy-based configuration.<br>3. <b>AIOps. </b>Machine learning (ML) with telemetry and decision support. In multivendor environments, vendors continuously analyze network behavior in real time.</p> <p>In the modern AIOps era, vendors use the following features for network management:</p> <ul class="default-list"> <li>ML for anomaly detection.</li> <li>Correlation engine for incident grouping.</li> <li>Stream telemetry processing.</li> <li><a href="https://www.techtarget.com/searchnetworking/definition/What-are-predictive-networks-and-how-do-they-work">Predictive modeling</a> for performance trends.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="10 network tasks AIOps can automate"> <h2 class="section-title"><i class="icon" data-icon="1"></i>10 network tasks AIOps can automate</h2> <p>AIOps environments already have several functions operating. These tasks are autonomous, meaning they execute continuously without human intervention once configured<b>.</b></p> <ol class="default-list"> <li><b>Anomaly detection. </b>AI is an intelligent layer. It continuously learns normal network behavior and flags deviations, such as latency spikes, device instability and traffic surges. It relies on ML models rather than fixed rules.</li> <li><b>E</b><b>vent c</b><b>orrelation and</b><b> n</b><b>oise r</b><b>eduction.</b> Many events happen on a network. AI groups thousands of raw alerts into a single meaningful incident by identifying relationships between events across devices, services and applications.</li> <li><b>Fault detection. </b>The system can automatically detect failures in links, routers, switches or services by <a href="https://www.techtarget.com/searchnetworking/tip/End-to-end-network-observability-for-AI-workloads">using telemetry signals</a> such as packet loss and interface status.</li> <li><b>Telemetry processing and analysis. </b>AI handles and processes massive streams of logs, metrics and network flows in real time. It can identify patterns and operational signals that humans can't analyze manually at scale.</li> <li><b>Network inventory discovery.</b> As an intelligent layer, AI can automatically detect devices connected to the network, map topology changes and maintain a current inventory without manual intervention.</li> <li><b>Application performance monitoring. </b>AI can learn normal network behavior and can predict potential issues before users report them. It <a href="https://www.techtarget.com/searchenterprisedesktop/definition/Application-monitoring-app-monitoring">tracks application behavior</a> by analyzing latency, jitter and throughput.</li> <li><b>Security anomaly detection. </b>AI can identify unusual network traffic behavior, such as <a href="https://www.techtarget.com/searchsecurity/feature/How-AI-threat-detection-is-transforming-enterprise-cybersecurity">unauthorized access</a>, abnormal traffic flows and excessive server requests.</li> <li><b>Incident ticket automation. </b>AI is an assistant to network teams, not a competitor. When issues are detected, it can automatically create and update tickets in <a href="https://www.techtarget.com/searchitoperations/definition/ITSM">IT service management</a> systems with relevant context, logs and severity classification without manual input from security teams.</li> <li><b>Dynamic baseline generation. </b>Instead of fixed limits, AI continuously updates what it considers normal for each device, app or network based on how traffic changes over time.</li> <li><b>Alert prioritization. </b>AI ranks incidents by severity and potential business consequences, surfacing critical issues first without manual triage.</li> </ol> </section> <section class="section main-article-chapter" data-menu-title="AIOps implications for network teams"> <h2 class="section-title"><i class="icon" data-icon="1"></i>AIOps implications for network teams</h2> <p>AI is shifting the networking operating model. With network functions being increasingly automated, engineers spend less time on manual troubleshooting. This means network teams can focus on other areas of network management, such as system design and governance.</p> <p>Networking professionals have shifted focus to the following new responsibilities:</p> <ul class="default-list"> <li>Automation policy definitions.</li> <li>Observability architecture design.</li> <li>AI-generated insight validation.</li> <li>AI system data quality management.</li> </ul> <p>This shift is real, but gradual. In some companies, legacy operations remain in place and take time to phase out. However, regardless of the type of intelligence AI brings into networking, a trust-but-verify mindset remains crucial, as AI can hallucinate and lead to erroneous decisions.</p> </section> <section class="section main-article-chapter" data-menu-title="Best practices for autonomous network operations"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Best practices for autonomous network operations</h2> <p>Organizations can adopt the following best practices to implement AIOps successfully within network operations:</p> <ol class="default-list"> <li><b>Build telemetry-first architectures. </b>AI success depends on data quality and coverage.</li> <li><b>Automate low-risk domains first. </b>Start with detection and correlation before moving to the control layers.</li> <li><b>Maintain human governance over changes</b>.<b> </b>Autonomy without control can increase risks, even as it reduces toil.</li> <li><b>Upskill engineering teams. </b>To adapt to industry disruption, network engineers should understand automation pipelines alongside more fundamental networking skills, such as routing protocols.</li> </ol> <p>Autonomous networking doesn't replace network engineers -- it removes the need to manually analyze network data at scale, giving engineers more time to focus on critical tasks where human judgment remains essential.</p> <p>Today, AI can detect issues, group alerts, prioritize issues and analyze telemetry. It can't, however, make high-risk changes to the network. Understanding this boundary is crucial for enterprises that want to use AI without creating organizational chaos.</p> <p><em>Verlaine Muhungu is a self-taught tech enthusiast, DevNet advocate and aspiring Cisco Press author, focused on network automation, penetration testing and secure coding practices. He was recognized as a Cisco top talent in sub-Saharan Africa during the 2016 NetRiders IT Skills Competition.</em></p> </section> AI can automate key network operations tasks, such as anomaly detection, event correlation and ticketing. This shifts network engineers toward governance and system design. https://cdn.ttgtmedia.com/rms/onlineimages/iot_a253400028.jpg https://www.techtarget.com/searchnetworking/tip/AI-driven-network-management-tasks Fri, 08 May 2026 17:15:00 GMT <p>Most organizations have embraced zero trust, but many are early in their adoption journey. Yet with the rising volume, velocity and sophistication of attacks, security teams are under pressure to accelerate those journeys.</p> <p>"We're definitely seeing higher rates of adoption today than one or two years ago," said Jimmy Nilsson, vice president of professional services at Kyndryl, a security consulting firm.</p> <p>Zscaler's ThreatLabz 2026 VPN Risk Report found that 84% of surveyed organizations had or were planning to implement a zero trust, up from 81% the prior year and 78% the year before that.</p> <p>Those figures, however, tell only part of the story. Researchers, security advisers and others in the field say enterprise security teams have just begun to take advantage of what zero trust can do to counter the many threats they face.</p> <p>Let's examine what zero trust is capable of and the specific uses cases where it can be put to work.</p> <section class="section main-article-chapter" data-menu-title="Zero-trust's capabilities"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Zero-trust's capabilities</h2> <p>Cybersecurity professionals view <a href="https://www.techtarget.com/searchsecurity/definition/zero-trust-model-zero-trust-network">zero trust</a> as an approach, a framework, a philosophy and a security model. Mike Monday, managing director of security and privacy at global business consulting firm Protiviti, called it an "engineering strategy."</p> <p>Zero trust is built on the idea that no user, device, system, workload or network segment -- even if it sits within an enterprise perimeter -- should be inherently trusted. Instead, the zero-trust security model requires entities to be authenticated and verified before they can access resources. Every access request must be authenticated, authorized and continuously validated based on identity, device health, context and risk signals.</p> <p>"That whole authentication has to happen through that end-to-end process," Monday explained.</p> <p>By removing inherent trust and adding authentication requirements and continuous validation, zero trust helps ensure that only authorized, authenticated entities are permitted access to an organization's IT environment and the data it holds. It also helps contain entities that do gain access, such as threat actors, by preventing unauthorized entities from moving freely throughout the environment.</p> <p>John Kindervag <a href="https://www.techtarget.com/whatis/feature/History-and-evolution-of-zero-trust-security">introduced the zero-trust security model</a> in 2010 while he was an analyst at Forrester Research. He and other early advocates championed zero trust as a necessary replacement for the traditional castle-and-moat security model, which by default extends trust to anything within the corporate environment. Such a hard-perimeter, soft-interior model relies on firewalls. In an era when cloud computing and other technologies were quickly eliminating the perimeter, <a href="https://www.techtarget.com/searchsecurity/tip/Perimeter-security-vs-zero-trust-Its-time-to-make-the-move">this approach provided inadequate protection</a> against threat actors.</p> <p>A zero-trust environment requires a combination of security technologies and IT architecture patterns and principles. These technologies include identity and access management, <a href="https://www.techtarget.com/searchsecurity/definition/multifactor-authentication-MFA">MFA</a>, zero trust network access (<a href="https://www.techtarget.com/searchnetworking/tip/The-basics-of-zero-trust-network-access-explained">ZTNA</a>) and endpoint detection and response tools. Key enabling IT architectures include microsegmentation and microperimeters.</p> <p>"Zero trust is a journey. It's a way of leveraging various technologies to address a specific problem, which is securing networks and securing data," said Fritz Jean-Louis, principal cybersecurity advisor at Info-Tech Research Group.</p> </section> <section class="section main-article-chapter" data-menu-title="Key use cases for zero trust"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Key use cases for zero trust</h2> <p>An organization can apply zero-trust principles in a variety of ways. Key use cases include the following:</p> <ul type="disc" class="default-list"> <li><b>Employees working on-site</b><b>.</b> Zero trust ensures on-site workers access only the systems and data necessary to perform their jobs at the time they need that access. This limits the <a href="https://www.techtarget.com/searchsecurity/tip/Five-common-insider-threats-and-how-to-mitigate-them">risks posed by insider threats</a>.</li> <li><b>Remote workers</b><b>.</b> With zero trust, remote workers access only the systems and data they are authorized to access when that access is required. They do so from devices and networks that are secure through contextual security enabled by ZTNA and other measures.</li> <li><b>Third parties</b><b>.</b> Zero trust can be applied to third parties outside the organization, such as contractors, partners and customers. Strictly controlled access for them reduces the risk of unwanted, unintended exposure and third party-related data breaches.</li> <li><b>System-to-system or machine-to-machine access.</b> These require continuous authentication for every request, and zero trust adds protection through the <a href="https://www.techtarget.com/searchsecurity/tip/Why-zero-trust-requires-microsegmentation">use of microsegmentation</a>. This zero-trust use case helps <a href="https://www.techtarget.com/searchsecurity/answer/Use-microsegmentation-to-mitigate-lateral-attacks">prevent lateral movement</a> by entities and ensures that if one service or device is compromised, attackers cannot automatically access other parts of the environment.</li> <li><b>Endpoints and remote devices.</b> In this use case, which includes operational and IoT technologies, zero trust requires that devices be authenticated and validated before they are permitted to access networks, systems and data.</li> <li><b>Access to APIs</b><b>.</b> Zero trust can be used for strict, continuous authentication and authorization for every API request, regardless of origin. This design is meant to permit legitimate access while preventing lateral movement by unauthorized entities. The result is a minimized blast radius in the event of an unauthorized entry somewhere in the environment.</li> <li><b>Data.</b> Zero trust can help protect data in the era of generative AI and large language models by authenticating and verifying AI identities and roles before granting them access to data they are authorized to use. Gartner has predicted that 50% of organizations will implement a zero-trust posture for data governance by 2028. This is increasingly relevant as unverified AI-generated data proliferates.</li> <li><b>AI agents</b><b>.</b> Organizations that <a href="https://www.techtarget.com/searchsecurity/tip/How-to-implement-zero-trust-for-AI">apply zero trust to AI agents</a> deny trust by default. Instead, agents are assigned individual identities, which enables each to be tracked. Zero trust prevents agents from sharing credentials, and agents are subject to continuous authentication and task-based permissions, as well as behavioral and semantic analysis.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Zero-trust implementation strategies and challenges"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Zero-trust implementation strategies and challenges</h2> <p>To <a href="https://www.techtarget.com/searchsecurity/feature/How-to-implement-zero-trust-security-from-people-who-did-it">implement or advance their use of zero trust</a>, experts advise organizations to develop a new mindset, yet many struggle to do this, Nilsson said.</p> <p>"Many organizations aren't successful because they're too focused on cybersecurity technology. They end up with siloed cybersecurity technologies, which is no different than how security organizations focused on cybersecurity two decades ago," Nilsson said. "Zero trust requires a new operating model. It's a change in how organizations approach security architecture."</p> <p>Nilsson and others cautioned organizations against implementing zero trust in every area of their digital environment all at once.</p> <p>"Zero trust can protect the entire ecosystem, but realistically, the number of tools you'd have to deploy to protect all those elements is onerous," Jean-Louis said.</p> <p>Experts also noted that organizations might struggle to implement zero-trust principles in legacy systems and to balance user experience with zero-trust requirements.</p> <p>Jean-Louis said he advises organizations to identify their <i>protect surface</i> -- that is, the portion of the larger attack surface they deem most necessary to protect. Consider how to apply zero trust to identities, devices, applications, data and the network using tools and technologies that can work across as many of those five areas as possible.</p> <p>Nilsson recommended a similar strategy, saying organizations should be as specific as they can in how they define their use cases. Build a zero-trust strategy for a specific use case, he said, and then use that as a blueprint for the next use cases.</p> <p>"Always think about what you are trying to secure, understand the asset you're trying to secure, how it is used by the business, how it collaborates with other systems in the business, and then build the security around that," Nilsson said.</p> <p><i>Mary K. Pratt is an award-winning freelance journalist with a focus on covering enterprise IT and cybersecurity management.</i></p> <p></p> </section> When applied correctly, zero trust can minimize an organization's attack surface. Experts weigh in on the best use cases where zero trust can deliver results. https://cdn.ttgtmedia.com/visuals/German/zero-trust-adobe.jpg https://www.techtarget.com/searchsecurity/tip/Zero-trust-use-cases-highlight-both-its-benefits-and-misconceptions Fri, 01 May 2026 00:00:00 GMT <p>Simple Network Management Protocol and Common Management Information Protocol were both designed to provide a method to manage networks, but their origins differ.</p> <p>SNMP and CMIP were each released in the late 1980s. <a href="https://searchnetworking.techtarget.com/definition/SNMP">SNMP</a>, an application-layer protocol, was designed by the individuals who worked to develop the internet -- i.e., the <a href="https://searchnetworking.techtarget.com/definition/TCP-IP">TCP/IP</a> family of protocols. CMIP was designed by committees of the Open Systems Interconnection (<a href="https://searchnetworking.techtarget.com/definition/OSI">OSI</a>) organization.</p> <section class="section main-article-chapter" data-menu-title="SNMP vs. CMIP: Different goals"> <h2 class="section-title"><i class="icon" data-icon="1"></i>SNMP vs. CMIP: Different goals</h2> <p>SNMP and CMIP reflect the differing development philosophies of the internet and OSI communities. The internet community deemed simplicity as most important, especially in the early days when devices had limited memory and processor capacity. The OSI community, however, believed each protocol must be designed to be extremely flexible and to initially address future requirements.</p> <p>When SNMP and CMIP were being developed in the 1980s -- and continuing into the 1990s -- individuals in the two communities believed they were in competition. Many in the OSI organizations said internet protocols, like SNMP, would collapse under load, while internet developers often felt OSI's protocols were too complicated, difficult to implement and expensive in terms of memory and processor resources.</p> </section> <section class="section main-article-chapter" data-menu-title="SNMP"> <h2 class="section-title"><i class="icon" data-icon="1"></i>SNMP</h2> <p>The goal of SNMP developers was to provide a <a href="https://searchitoperations.techtarget.com/tip/SNMP-monitoring-tools-improve-how-IT-looks-at-application-performance">simple but useful protocol</a>. SNMP initially had just two commands: <span style="font-family: 'courier new', courier, monospace;">GET</span> and <span style="font-family: 'courier new', courier, monospace;">SET</span>. SNMP managers use these two commands to access and modify device parameters.</p> <p>The application-layer protocol initially had no security, but developers added security in later revisions. Developers also designed additional commands -- such as <span style="font-family: 'courier new', courier, monospace;">GETBULK</span> and <span style="font-family: 'courier new', courier, monospace;"> RESPONSE </span>-- to access more than one parameter at a time and for a device to signal the SNMP manager that a problem had occurred.</p> <p>Exchanges between management stations and managed devices are straightforward. They simply exchange datagrams that contain commands and responses. No connection between manager and device is maintained with SNMP.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineImages/networking-snmp_vs_cmip-f.png"> <img data-src="https://www.techtarget.com/rms/onlineImages/networking-snmp_vs_cmip-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineImages/networking-snmp_vs_cmip-f_mobile.png 960w,https://www.techtarget.com/rms/onlineImages/networking-snmp_vs_cmip-f.png 1280w" alt="SNMP vs. CMIP" height="272" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Compare the differences of SNMP vs. CMIP, including protocol suite, goals and exchange method. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="CMIP"> <h2 class="section-title"><i class="icon" data-icon="1"></i>CMIP</h2> <p>OSI developers believed a more flexible and comprehensive management protocol was needed as they considered SNMP vs. CMIP. In addition to retrieving and modifying network parameters, a CMIP manager can direct a remote application to create or delete a software entity that can then be managed.</p> <p>CMIP management stations maintain connections to devices using the <a target="_blank" href="https://tools.ietf.org/html/rfc1189" rel="noopener">Remote Operations Service Element</a> (ROSE) protocol to communicate with devices. ROSE is the OSI remote procedure call protocol used to carry CMIP commands between management stations and managed devices. All parameters are encoded using Abstract Syntax Notation One, an interface description language. Security capabilities keep hackers from modifying configurations.</p> <p>Although the OSI protocols are used for some applications today, the success of the internet has proven that fundamental communications protocols, including SNMP, can carry today's traffic -- a load far beyond anything imaginable in the 1980s and 1990s.</p> <p><b>Editor's note:</b><i>&nbsp;This article was updated in 2026 to improve the reader experience.</i></p> <p><em>David Jacobs has more than 30 years of networking industry experience. He has managed leading-edge hardware and software development projects as well as consulted Fortune 500 companies and software startups.</em></p> </section> While SNMP and CMIP originally competed as network management protocols in the 1980s, SNMP has emerged as the more relevant protocol, largely due to its simplicity. https://cdn.ttgtmedia.com/visuals/searchSMBStorage/data_storage_management/smbstorage_article_013.jpg https://www.techtarget.com/searchnetworking/answer/SNMP-vs-CMIP-Whats-the-difference Thu, 30 Apr 2026 16:30:00 GMT <p>Network analytics can be a rich source of intelligence about infrastructure performance, efficiency, security and operations. Given enterprise reliance on connectivity for everything from internal communications and customer service to production, analyzing network traffic from various sources can&nbsp;yield insights&nbsp;into key patterns and trends.</p> <p>As a result, network analytics use cases can illustrate how organizations have been able to troubleshoot configuration issues and improve network efficiency. IT professionals can also apply network analytics to cut operational costs and identify potential security threats.</p> <section class="section main-article-chapter" data-menu-title="3 essential network analytics use cases"> <h2 class="section-title"><i class="icon" data-icon="1"></i>3 essential network analytics use cases</h2> <p>Deriving the maximum benefit from network analytics requires IT professionals to <a href="https://www.techtarget.com/searchnetworking/answer/Wireshark-vs-tcpdump-Whats-the-difference">capture data from disparate sources</a> across the infrastructure and then correlate it to gain an end-to-end view of patterns. This process requires merging data from different systems. While this is often complicated, it enables network professionals to address important network analytics use cases, such as the following:</p> <ul class="default-list"> <li>Performance optimization and capacity planning.&nbsp;</li> <li>Credential misuse.&nbsp;</li> <li>Cloud security.&nbsp;</li> </ul> <h3>Performance optimization and capacity planning</h3> <p>When done effectively, network analytics reveals crucial information about hidden bottlenecks and other&nbsp;<a href="https://www.techtarget.com/searchnetworking/tip/Network-design-principles-for-effective-architectures">network design&nbsp;</a>issues that can choke traffic and impede productivity. To illustrate one of the benefits of network analytics use cases, companies find that applying that data to gauge ongoing performance optimization could return even greater dividends.</p> <p>IT organizations can use historical pattern information to anticipate future capacity requirements and potential performance issues. By tapping into this data, IT organizations can rebalance network loads, expand capacity and make configuration changes to better adjust to communication requirements.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/qNvzbvC_bm8?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <h3>Credential misuse</h3> <p>Network analytics can also shed light on security issues, applying network telemetry data alongside other data sources to accelerate the discovery of anomalous patterns indicative of threats.&nbsp;<a href="https://www.techtarget.com/searchdatamanagement/tip/Residual-access-failures-put-data-at-risk">Credential misuse</a>, which is often difficult to detect through manual human analysis, can be spotted much more quickly when comparing abnormal behavior against a normal baseline.</p> <p>Network analytics can flag unusual actions, such as multiple login tries from different devices or attempts to connect to a network resource from an unexpected device.</p> <p>If a&nbsp;breach has occurred, IT professionals can use network analytics to expedite the discovery of which assets the rogue user was granted access to and what data could be compromised. Expedited breach identification can help <a href="https://www.techtarget.com/searchsecurity/feature/How-to-develop-a-data-breach-response-plan-5-steps">prevent future losses</a> and mitigate the consequences of the theft.</p> <h3>Cloud security</h3> <p>IT organizations can also use flow logs from third-party cloud environments to get more insight into activity in their&nbsp;on-demand environments, a good example of network analytics use cases. <a target="_blank" href="https://medium.com/@samruddhi.deshmukh19/cloud-analytics-6a06569e608f" rel="noopener">Network analytics from the cloud</a> can help organizations track performance and make more accurate capacity-planning decisions for commissioning and decommissioning of cloud resources based on dynamic usage requirements. Cloud network analytics can also shed light on potential security threats or possible vulnerabilities.</p> </section> <section class="section main-article-chapter" data-menu-title="The value of network analytics use cases"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The value of network analytics use cases</h2> <p>It's important for enterprises to recognize use cases for network analytics tools. Use cases not only demonstrate the value of network analytics but also show where and how the tools excel. Network analytics help organizations evaluate network performance and identify anomalies. Moreover, when the network performs as expected, organizations can gather information about its behavior to optimize performance, improve troubleshooting and strengthen security.</p> <p><b>Editor's note:</b><i>&nbsp;This article was updated in 2026 to improve the reader experience.</i></p> <p><em>Amy Larsen DeCarlo has covered the IT industry for more than 30 years, as a journalist, editor and analyst. As a principal analyst at GlobalData, she covers managed security and cloud services.</em></p> </section> Network analytics enables teams to monitor performance, identify issues early and improve efficiency. These use cases illustrate how insights can address network challenges. https://cdn.ttgtmedia.com/visuals/IoTAgenda/business_of_iot/iotagenda_article_004.jpg https://www.techtarget.com/searchnetworking/answer/What-are-the-top-3-network-analytics-use-cases Thu, 30 Apr 2026 12:15:00 GMT <p>Cisco is working to extend its networking leadership into the age of quantum computing.</p> <p>The company revealed a prototype Thursday of what it calls a "universal quantum switch," one that understands the encoding modalities of different types of quantum computers so they can be networked in the same way classical computers are today.</p> <p>The switch works at room temperature and can hook into existing fiber optic networks. Unlike today's commercial <a href="https://www.techtarget.com/searchnetworking/news/252454768/Cisco-sees-the-future-of-switching-in-Luxtera-silicon-photonics">photonic switches</a>, these switches preserve the quantum entanglement of photons, allowing them to convey complex quantum states.</p> <p>"Cisco's quantum network greatly expands the possibilities of quantum computing by allowing the networking of several smaller quantum computers into an effectively larger one, even if the underlying physics of the quantum computers are substantially different," said Mark Horvath, a Gartner analyst, in an email.</p> <p>In February, Cisco, in partnership with Qunnect, <a target="_blank" href="https://outshift.cisco.com/blog/quantum/scalable-quantum-entanglement-networking" rel="noopener">set up</a> a quantum network over a noisy commercial fiber optic conduit from Brooklyn to the 60 Hudson Street telecommunication carrier hotel in Manhattan. They were able to convey 5,400 qubit pairs per hour over the 17.6-kilometer route, about 10,000 times faster than any previous experiments, Cisco researchers asserted.</p> <p>"Cisco Quantum's recent validation results from working with Qunnect in NYC were fascinating and good evidence of solid advancements, proving quantum networking can work over distances and without supercooling," said Jim Frey, Omdia principal analyst for networking, in an email.</p> <section class="section main-article-chapter" data-menu-title="A need for quantum networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>A need for quantum networking</h2> <p>Quantum computing promises to solve complex problems too difficult for today's classical computers, in areas such as financial modeling, drug discovery, weather exploration and logistics planning.</p> <p>Even after decades of research, however, quantum computers are still far from where they need to be to run these complex workloads, which would require hundreds of thousands or even millions of quantum bits -- or <a href="https://www.techtarget.com/whatis/definition/qubit">qubits</a> -- to model.</p> <p>Today's most powerful quantum computers, however, can juggle at most only a few thousand qubits at once, and with vigorous research in the next few years, that number might only jump to the low tens of thousands.</p> <p>"There's a big gap between the number of qubits we need and the number of qubits we have," said Vijoy Pandey, senior vice president and general manager of Outshift, Cisco's emerging technologies and incubation group, during a press conference this week.</p> <p>So, it would be natural to borrow a technique from the high-performance computing community and lash quantum computers together to get the required power needed to tackle these jobs in a distributed manner.</p> <p>A network of computers would scale faster than trying to vertically build up a single computer, Pandey argued.</p> <p>"In order to reach the total compute capacity needed for quantum computing to be broadly viable, and to make the system secure, it will be necessary to chain quantum computers together via quantum networking," Frey agreed.</p> </section> <section class="section main-article-chapter" data-menu-title="Cisco doing what it does best"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Cisco doing what it does best</h2> <p>Cisco's universal quantum switch works just like a typical data center switch but adjusted for quantum traffic. It accepts the incoming quantum signal, internally converts it to a neutral common modality for routing, and sends it out in the correct modality for the receiving system -- preserving the quantum state all the while.</p> <p>Classical internet routing differs from what would be required for the quantum realm. The internet operates on a store-and-forward approach, but a quantum network must establish an open path between the source and destination first to share the entanglement state.</p> <p>"For now, we think of quantum networking as completely different from classical networking. Distributing quantum states is a completely separate way of doing things," said Ramana Kompella, Cisco fellow and head of Cisco research, during the press conference.</p> <p>Quantum computing states are fragile, Kompella said. A pair of photons can be entangled to produce a qubit, and the more photons you can entangle, the more information you can convey. Moving a quantum state from one processor to another, called <i>teleportation</i>, requires preserving each qubit's entanglement properties. But electromechanical noise can easily destroy a qubit. The more devices a state of qubits passes through, the more a potential loss of fidelity compounds.</p> <p>Different quantum systems have different encoding modalities, much like early classical computers were incompatible with one another. <a href="https://www.techtarget.com/searchdatacenter/feature/Companies-building-quantum-computers">IBM quantum systems</a> have a different modality, for instance, from those built by Atom Computing. This is not vendor lock-in for its own sake, Kompella said. Different applications favor different modalities.</p> <p>Cisco partnered with both IBM and Atom for this switch, which will support four modalities:</p> <ul class="default-list"> <li>Polarization, where the orientation of light waves carries information.</li> <li>Time-bin, where the timing of light pulses carries information.</li> <li>Frequency-bin, where the color or frequency of light is used to carry information.</li> <li>Path, where the physical or spatial path carries information.</li> </ul> <p>"We don't know which of these technologies will be dominant," said Reza Nejabati, head of Cisco quantum research and quantum labs, during the press conference. "So the network can support any of them."</p> <p>Fast switching was also imperative. The Cisco switch can reconfigure connection in as little as a nanosecond, while consuming less than a milliwatt of power.</p> <p>Key to the switching process is a silicon-based quantum state converter, which, using quantum tomography, converts qubits into an internal encoding format for routing. They are then converted back, without being measured or destroyed, to their original encoding, or to one of the other encoding formats, at their destination.</p> <p>Unlike other qubit management approaches, the switch can operate at room temperature, thanks to the frequency of telecommunications fiber optic channels and the internal encoding format.</p> <p>The work builds on previous Cisco research as part of its Outshift quantum incubation program, including a quantum network entanglement chip to link the quantum computers to the network, and a network-aware quantum compiler that orchestrates quantum algorithms across multiple quantum processors -- handy for error correction.</p> </section> <section class="section main-article-chapter" data-menu-title="The next steps for quantum networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The next steps for quantum networking</h2> <p>The Cisco universal switch is still in the prototype phase. It will be several years before it is released commercially, as the supporting stack for end-to-end operation needs to be finalized. For instance, the company will need to build a repeater of some sort because quantum signals can only travel about 100 kilometers or so.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> This is a seismic breakthrough that will bring solutions into reach for many problems that were years beyond today's existing generation of quantum computers. </figure> <figcaption> <strong>Mark Horvath</strong>Gartner analyst </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>In the more immediate term, researchers will post a paper explaining the technology on ArXiv.</p> <p>"This is a seismic breakthrough that will bring solutions into reach for many problems that were years beyond today's existing generation of quantum computers," Horvath said.</p> <p>In addition to putting quantum computing on a more solid commercial footing, quantum networking may open up new uses on its own, Kompella said.</p> <p>For instance, a quantum network could give high-frequency trading (HFT) firms a competitive edge. An HFT firm might run across multiple data centers, and when it makes a purchasing decision in one data center, it should replicate as quickly as possible in the other locations.</p> <p>Today, the network is the bottleneck for HFT, in that it limits the speed at which that information can be moved to near the speed of light. <a href="https://www.techtarget.com/searchnetworking/tip/An-introduction-to-quantum-networks-and-how-they-work">Quantum communications</a> could beat what was previously thought to be the physical upper limit by coordinating the decision-making process ahead of time through entanglement, according to Pandey.</p> <p>"Coordinated decisions [that run] faster than the speed of light actually give you a substantial advantage compared to any classical strategy," Kompella said.</p> <p><i>Freelance news writer Joab Jackson has been writing about back-end IT technologies for the past three decades. His grandfather programmed mainframes, and his father wrote computer games for hobbyist programming magazines in the 1980s.</i></p> </section> Cisco's quantum switch prototype operates like a data center switch and supports four encoding modalities. It marks a step forward in making quantum computing 'broadly viable.' https://cdn.ttgtmedia.com/rms/onlineimages/arvr_g1273484747.jpg https://www.techtarget.com/searchnetworking/news/366642172/Cisco-unveils-quantum-network-advancements Thu, 23 Apr 2026 09:00:00 GMT <p>Contact center fraud is a reality that organizations must prepare for or else risk considerable losses due to security lapses in customer data protection. Successful fraud schemes can damage a brand's reputation and result in compliance liability, especially in heavily regulated industries, such as financial services and healthcare.</p> <p>As contact centers expand into digital channels and remote operations, fraud detection has become a critical component of customer experience and data security strategies.</p> <p>Companies can mitigate their vulnerability to unauthorized access or disclosure of confidential information with the right blend of <a href="https://www.techtarget.com/searchcustomerexperience/tip/Best-practices-for-call-center-agent-training-programs">comprehensive agent training</a>, well-documented authentication and data security processes, and contact center fraud detection technologies.</p> <section class="section main-article-chapter" data-menu-title="What is contact center fraud?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is contact center fraud?</h2> <p>At many businesses, traditional call centers and customer service and support operations have <a href="https://www.techtarget.com/searchcustomerexperience/feature/History-and-evolution-of-contact-centers">evolved into contact centers</a> to handle customer communications across multiple channels, including phone calls, live chats, email, social media, text messaging (SMS), mobile apps and video calls.</p> <p>Cybercriminals target contact centers to gain access to sensitive customer information by exploiting agents and weak authentication processes. These bad actors can then use personally identifiable information (<a href="https://www.techtarget.com/searchsecurity/definition/personally-identifiable-information-PII">PII</a>) and other account data -- Social Security numbers, financial institutions and credit card numbers -- to commit identity theft, set up fake accounts and participate in bank and credit card fraud.</p> </section> <section class="section main-article-chapter" data-menu-title="Why do bad actors target contact centers?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Why do bad actors target contact centers?</h2> <p>Contact centers are popular targets for fraud because poorly trained agents are often vulnerable to manipulation. A toll-free number used for customer service and transactions such as purchases can allow criminals to initiate numerous fraud attempts while maintaining anonymity, provided they use caller ID spoofing techniques. Unsuspecting agents, especially in call centers, make excellent attack vectors since they're all that stand between a fraudster and customer accounts.</p> <p>The expansion of <a href="https://www.techtarget.com/searchcustomerexperience/tip/How-to-manage-remote-call-center-agents">hybrid and remote contact center operations</a> has introduced new fraud detection challenges. Remote work has made it increasingly difficult for agents to receive proper fraud detection training or guidance from co-workers. As a result, they may struggle with using anti-fraud tools remotely.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f.png 1280w" alt="Graphic showing a contact center compliance checklist, including securing networks, authenticating customers, recording conversations and managing sensitive information." height="266" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Contact center compliance programs help organizations reduce fraud risk by securing networks, authenticating customers, protecting sensitive data and following privacy and consumer protection regulations. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Common types of contact center fraud"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Common types of contact center fraud</h2> <p>While contact centers encounter many types of fraud, the most common are identity theft, account takeover, stolen credit card information and finagling free merchandise.</p> <p><b>Identity theft.</b> Criminals use stolen personal information of legitimate customers to access accounts for monetary gain. Contact center agents might struggle to detect identity theft because the bad actors have accurate customer information. Many fraud schemes use personal information found on the dark web after a data breach. Synthetic identity fraud occurs when criminals combine real PII, such as a mobile phone number and email address, with falsified data to create a manipulated or false identity. They then use the information to open accounts and initiate transactions.</p> <p><b>Account takeover.</b> To transfer a customer account to their account, fraudsters might change an email address or login information to reset customer portal passwords. These criminals can use automated tools to create username and password combinations in a technique known as credential stuffing to gain access to customer accounts.</p> <p><b>Use of stolen credit card information.</b> Fraudsters bombard contact centers with attempts to buy goods and services with stolen credit card information. Because contact centers don't require physical cards, criminals can more easily make purchases with stolen information, a tactic known as card-not-present fraud.</p> <p><b>Attempt to receive free replacement items.</b> Criminals act as legitimate customers who purchased goods, then claim to have problems and request replacements. Retailers are the most common victims of this type of fraud, especially those with loose warranty and replacement policies.</p> <p><b>Phishing and vishing scams. </b>Cybercriminals have long targeted consumers with phishing scams, sending fraudulent emails that contain malicious URLs or hyperlinks to download malware or steal passwords. Another tactic is <i>voice phishing</i>, or <i>vishing</i>, using urgent phone calls that demand victims to update company or personal data supposedly to protect bank accounts and other financial transactions. Similar fraudulent methods are used on contact center agents. A criminal vishing about problems with an account can dupe an unsuspecting agent into sharing sensitive customer data.</p> <p>Many contact centers have been hit with ransomware attacks, locking up communications systems until the problem is resolved or the ransom is paid. <a href="https://www.techtarget.com/searchsecurity/definition/distributed-denial-of-service-attack">Distributed denial-of-service attacks</a> have also been used to disrupt communications services. More recently, AI-generated voice cloning and deepfake audio can be used to impersonate legitimate customers.</p> </section> <section class="section main-article-chapter" data-menu-title="Tips for identifying fraudulent customers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Tips for identifying fraudulent customers</h2> <p>Criminals use different fraud methods depending on their motivation or the <a href="https://www.techtarget.com/searchcustomerexperience/feature/Types-of-contact-centers-explained">type of contact center</a> they target. Common warning signs of fraud include the following:</p> <ul type="disc" class="default-list"> <li>Social engineering methods to falsely extract information.</li> <li>Inability to verify recent transactions.</li> <li>Long pauses before answering questions.</li> <li>Communication to evoke an immediate reaction based on urgency, familiarity or authority.</li> <li>Attempts to establish a relationship or rapport with a specific contact center agent or manager.</li> <li>Inconsistency in customer history and documentation.</li> <li>Attempts to bypass regular customer service procedures.</li> <li>Red flags and suspicious activity identified by anti-fraud technologies.</li> <li>Attempts to bypass anti-fraud processes and technologies.</li> <li>Automated speech patterns that may indicate AI-generated voice fraud.</li> </ul> <ul class="default-list"></ul> </section> <section class="section main-article-chapter" data-menu-title="Tools to identify fraud"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Tools to identify fraud</h2> <p>Enterprises that take contact center fraud detection and prevention seriously shouldn't rely solely on agent training. Contact center managers can integrate several technologies into most on-premises, cloud or distributed workforce contact centers to block or flag suspicious activities and enhance fraud detection.</p> <p><b>Identity verification. </b>Technologies like automatic number identification can verify a customer's identity based on their phone number ahead of automated or interactive voice response (<a href="https://www.techtarget.com/searchcustomerexperience/definition/Interactive-Voice-Response-IVR">IVR</a>) interactions. Some of these fraud detection technologies track phone numbers based on information like possession (authenticating the mobile number and the device), reputation (risk score) and ownership. If additional verification is needed, layered authentication controls can help prevent fraud by sending one-time verification codes via text or email to a customer's device. In the future, individuals could have additional ways to prove their identity with mobile devices as more states offer digital driver's licenses and government IDs. Some identity verification platforms now combine device fingerprinting, behavioral analytics and risk scoring.</p> <p><b>Contact source analytics. </b>Emerging technologies can more accurately confirm a contact's true source as well as the type of device used. These attributes can tip off contact center agents about whether the caller is a real customer or a criminal in a known fraud location or using equipment common among fraudsters, such as caller ID spoofing and IVR probing tools.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/ai_sharpens_contact_center_features_and_actions-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/ai_sharpens_contact_center_features_and_actions-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/ai_sharpens_contact_center_features_and_actions-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/ai_sharpens_contact_center_features_and_actions-f.png 1280w" alt="Diagram showing how AI improves contact center features such as IVR systems, self-service chatbots, agent performance analytics and post-call summaries." height="355" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>AI technologies can strengthen contact center operations by improving IVR systems, enabling self-service chatbots, supporting agent performance monitoring and generating automated post-call summaries. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p><b>Multilayered authentication.</b> Multifactor authentication, AI and knowledge-based platforms can identify bad actors who impersonate legitimate customers. The technology platform inputs various data points and calculates a fraud risk score to inform the agent about next steps in the fraud prevention process. A one-time pin or passcode sent by text or email to an individual's device can add a dynamic layer of security before a login session or transaction. Based on risk assessments, businesses must find the right balance between frictionless customer experience and layered security measures.</p> <p><b>Voice biometrics.</b> Advanced audio biometrics can analyze a caller's voice, creating a new authentication layer for contact centers and customers. Voice biometric SaaS providers let remote agents access these authentication services regardless of where they work. These technologies will soon have to contend with AI-driven voice cloning and deepfake audio, which might require reevaluation of fraud protection and other security measures.</p> <p><b>Suspicious behavior detection.</b> <a href="https://www.techtarget.com/searchcustomerexperience/feature/Important-contact-center-AI-features-and-their-benefits">AI and machine learning techniques</a> combine with fraud detection analytics tools to detect suspicious behavior such as unusual calling patterns, IVR usage anomalies and other behavior-based indicators. The tool then decides whether the contact is legitimate. Behavioral analytics can also be used to monitor agent behavior for insider threats by flagging multiple account redirects or password resets.</p> <p>Organizations that combine agent training with layered authentication and AI-driven fraud detection tools are better positioned to protect customer data and maintain trust.</p> <p><b>Editor's note:</b> <i>This article was updated to reflect the latest developments in contact center fraud detection and prevention tools, techniques and practices.</i></p> <p><i>Kathleen Richards is a freelance journalist and industry veteran. She's a former features editor for TechTarget's </i>Information Security <i>magazine.</i></p> <p><i>Andrew Froehlich is founder of InfraMomentum, an enterprise IT research and analyst firm, and president of West Gate Networks, an IT consulting company. He has been involved in enterprise IT for more than 20 years.</i></p> </section> Scammers may target contact centers, but comprehensive agent training, authentication techniques and advanced technologies can protect businesses and customers. https://cdn.ttgtmedia.com/rms/onlineimages/customer_service02.jpg https://www.techtarget.com/searchcustomerexperience/tip/How-to-train-agents-on-call-center-fraud-detection Tue, 14 Apr 2026 11:02:00 GMT <p>Legacy networks, older infrastructures built on outdated hardware and OSes, have many limitations compared to modern networks and pose significant business risks.</p> <p>Despite their risks, legacy networks persist due to the following:</p> <ul class="default-list"> <li>Engineers rely on existing skill sets to maintain networks.</li> <li>Organizations don't have the <a href="https://www.techtarget.com/searchnetworking/tip/How-to-develop-a-network-budget">budget to allocate toward a network upgrade</a>.</li> <li>Executives don't recognize the importance of funding new purchases for the network.</li> </ul> <p>Neglecting to upgrade an outdated network, however, is a critical mistake organizations can't afford to make. The continued use of a legacy network creates management complexities, business risks and higher costs. This article examines the risks of legacy networks, the benefits of transitioning and best practices to get started.</p> <section class="section main-article-chapter" data-menu-title="The limitations of legacy networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The limitations of legacy networking</h2> <p>Many organizations still use legacy networks, but they come with limitations that prevent them from reaching their full potential. Examples of limitations include the following:</p> <ul class="default-list"> <li><b>Device management complexity</b>. Legacy networks use tools that parse human-readable command-line interface text. Because these legacy devices lack APIs, it's difficult to integrate them with modern automation platforms.</li> <li><b>Performance bottlenecks</b>. Legacy devices lack application-specific integrated circuit-based acceleration that is required for advanced networking features. As a result, AI workloads in high-performance environments might not have access to low-latency communication. Combined with low network interface card bandwidth, this creates a data transfer bottleneck.</li> <li><b>Legacy hardware</b>. Vendors often stop sending software updates to old devices. These devices use older <a href="https://www.techtarget.com/searchnetworking/feature/Wireless-encryption-basics-Understanding-WEP-WPA-and-WPA2">wireless encryption standards</a>, such as WPA2, because their hardware cannot support WPA3. These older standards lack newer features modern networks require.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Business risks of legacy networks"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Business risks of legacy networks</h2> <p>When legacy networks are used today, their limitations translate into real operational and business risks. These risks include the following:</p> <ul class="default-list"> <li><b>Vulnerable to attacks.</b> Outdated encryption and unpatched devices can expose a network to attacks. For example, Wi-Fi devices that still use WPA2 pose a significant risk when WPA3 is available in newer models.</li> <li><b>Loss of revenue</b>. Because legacy networks are more susceptible to attacks, these hacks can <a href="https://www.techtarget.com/searchnetworking/feature/The-effects-of-network-downtime-and-ways-to-fix-it">cause network downtime</a> and service disruption.</li> <li><b>Operational inefficiency from silos</b>. Virtual devices in public clouds and physical on-premises devices are two different silos. Manually configuring each of these silos increases the risk of human error.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="The cost of inaction"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The cost of inaction</h2> <p>Legacy networks can incur various financial and operational costs for organizations due to outdated infrastructure. Examples include the following:</p> <ul class="default-list"> <li><b>Modern applications limited by legacy networking</b>. If an organization tries to <a href="https://www.techtarget.com/searchcio/feature/How-to-modernize-legacy-applications">modernize applications</a> by adopting Kubernetes or using open source software tools, the disparity between the legacy network and the new technologies could present challenges. Examples include forced reliance on disparate load balancers, lack of zero-trust security, insufficient fine-grained traffic control and limited observability at the application level.</li> <li><b>Inability to automate.</b> Legacy networks have limited automation capabilities, which means network teams complete tasks manually. Teams will become slower to perform device upgrades or troubleshoot.</li> <li><b>Barriers to AI adoption</b>. Over time, the inability to use higher-performance networking infrastructure compounds into a structural disadvantage. Organizations will stagnate as their competitors continuously improve their services with AI.</li> <li><b>Competitive disadvantage</b>. The inability to use new networking technologies leads to the loss of <a target="_blank" href="https://rogerlmartin.substack.com/p/first-mover-advantages-versus-pioneering" rel="noopener">first-mover advantage</a>. Organizations could be outpaced by competitors in digital product releases.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Moving beyond legacy networks"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Moving beyond legacy networks</h2> <p>To successfully transition from legacy networking architectures, organizations should follow key best practices. Recommendations include the following:</p> <ul class="default-list"> <li><b>Assess and prioritize workloads</b>. Identify high-throughput and low-latency applications. This ensures the network team knows exactly how much capacity and performance the upgrades must deliver.</li> <li><b>Adopt automation and programmability</b>. Purchase devices that support APIs, and ask engineers to <a href="https://www.techtarget.com/searchnetworking/tip/Best-practices-for-network-automation-with-Python">learn Python to become familiar with network automation and AI</a>. If part of a workload runs in a public cloud, use orchestration platforms to automate configuration across both environments.</li> <li><b>Ensure security by design</b>. Purchase devices with modern encryption, such as WPA3. For application-level security, use mutual <a href="https://www.techtarget.com/searchsecurity/definition/Transport-Layer-Security-TLS">TLS</a> encryption in service meshes.</li> <li><b>Transition in phases</b>. A phased transition enables the organization to continue operating without network disruption.</li> </ul> <p>By following these recommendations, organizations can transition outdated legacy networks into modern networking architectures.</p> <p><em>Charles Uneze is a technical writer who specializes in cloud-native networking, Kubernetes and open source.</em></p> </section> Legacy networks face limitations like security vulnerabilities and performance issues, leading to operational and business risks. Transitioning to modern infrastructure is crucial. https://cdn.ttgtmedia.com/rms/onlineimages/disaster_recovery_g179288049.jpg https://www.techtarget.com/searchnetworking/tip/Why-legacy-networks-are-a-growing-liability Fri, 10 Apr 2026 17:00:00 GMT <p>Contact centers and their agents are a critical part of customer service. They're the ambassadors of the organization, responding to large call volumes daily, interacting with customers and collecting feedback to pass on to the business.</p> <p>Modern contact center platforms increasingly use AI-driven analytics, speech recognition and sentiment analysis tools to monitor interactions in real time and identify opportunities to improve both agent performance and customer experience (CX).</p> <p>A contact center monitoring program can help businesses <a href="https://www.techtarget.com/searchcustomerexperience/feature/The-ultimate-guide-to-contact-center-modernization">transition the contact center from an expense center to a strategic asset</a> by ensuring representatives effectively resolve customer issues along with&nbsp;capturing valuable customer feedback. Many companies have a basic QA monitoring program but often struggle with transitioning to a more advanced one. Businesses should identify the benefits of an advanced quality monitoring program and implement key best practices to ensure the program's success.</p> <section class="section main-article-chapter" data-menu-title="What is a contact center monitoring program?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is a contact center monitoring program?</h2> <p>A basic contact center quality monitoring program consists of listening to phone calls between customers and contact center agents and <a href="https://www.techtarget.com/searchcustomerexperience/answer/5-ways-to-improve-call-center-agent-performance">providing feedback to improve agent performance</a>.</p> <p>An advanced QA monitoring program adds three key elements:</p> <ol class="default-list"> <li>Provides insight into why customers call and facilitates action plans to address the root cause of customer inquiries.</li> <li>Identifies <a href="https://www.techtarget.com/searchcustomerexperience/tip/Contact-center-challenges-and-how-to-overcome-them">customers who are frustrated with the company</a> and might decide to do business with a competitor.</li> <li>Analyzes the tools that agents use and implements enhancements to those tools that improve the agent experience and provide more accurate and timely responses to customers.</li> </ol> <p>&nbsp;Many organizations now augment traditional QA monitoring programs with AI-driven analytics tools that automatically analyze call transcripts and customer sentiment.</p> </section> <section class="section main-article-chapter" data-menu-title="What are the benefits of contact center monitoring?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the benefits of contact center monitoring?</h2> <p>A contact center is the place where the voice of the customer is heard. It's the one place in the organization where a large number of customers reach out and, in most cases, provide unsolicited feedback to the company. A well-designed contact center monitoring program provides a valuable opportunity to identify customer pain points and gather intelligence with the goal of improving products, services and overall CX.</p> <p>Retaining existing customers is typically less expensive than acquiring new ones, so it's critical to <a href="https://www.techtarget.com/searchcustomerexperience/tip/How-to-improve-the-contact-center-experience-for-customers">identify areas for improvement in the current customer</a> base to increase retention and reduce costs. Contact center monitoring also provides real-time information at a much more granular level than either customer satisfaction or Net Promoter Score surveys, which are performed after the fact and have some level of bias, depending on who does or doesn't respond to a survey request.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineImages/crm-call_center_agent_characteristics.png"> <img data-src="https://www.techtarget.com/rms/onlineImages/crm-call_center_agent_characteristics_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineImages/crm-call_center_agent_characteristics_mobile.png 960w,https://www.techtarget.com/rms/onlineImages/crm-call_center_agent_characteristics.png 1280w" alt="Graphic listing key qualities of a contact center agent, including knowledgeable, detail-oriented, organized, flexible, empathetic and effective communicator." height="288" width="559"> <figcaption> <i class="icon pictures" data-icon="z"></i>Successful contact center agents combine interpersonal and organizational skills such as communication, empathy, flexibility and problem-solving to deliver strong customer service. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="How to start a contact center monitoring program"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to start a contact center monitoring program</h2> <p>Starting and developing a contact center monitoring program require several steps, including the following:</p> <ol class="default-list"> <li>Identify the criteria that is monitored and scored, such as greeting, tone, call documentation and adherence to procedures.</li> <li>Develop a scorecard that measures the items to be monitored.</li> <li>Determine who performs the monitoring, such as a supervisor or QA analyst.</li> <li>Set the frequency of monitoring per agent and when the monitoring occurs.</li> <li>Develop a process to provide feedback to agents.</li> <li>Let the agents know the purpose of the monitoring program and how it works.</li> <li>Test the quality monitoring process end to end.</li> </ol> </section> <section class="section main-article-chapter" data-menu-title="Contact center monitoring best practices"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Contact center monitoring best practices</h2> <p>Successful quality monitoring programs typically include the following best practices.</p> <h3>1. Define quality and the ideal customer interaction</h3> <p>Contact center agents can't provide the proper service to customers if they don't know what the company expects of them. So, it's important for <a href="https://www.techtarget.com/searchcustomerexperience/definition/contact-center-management">contact center management</a> to train employees on what to say and do during a customer interaction before beginning the monitoring process. Scripts for agents are sometimes a contact center practice and other times a legal requirement, but they can help agents start off on the right foot by giving them a roadmap of what to say and how an interaction should be done. When scripts aren't a legal requirement, it's often beneficial to modify and use them as a guideline and make them less robotic-sounding to better serve customers.</p> <h3>2. Decide what customer service metrics are most important</h3> <p>Businesses shouldn't try to measure everything. Contact center managers need to <a href="https://www.techtarget.com/searchcustomerexperience/tip/Top-7-call-center-agent-performance-metrics-to-track">decide what metrics they value the most</a> and communicate them to their teams before beginning the quality monitoring process. Some metrics include first-contact resolution (<a href="https://www.techtarget.com/searchcustomerexperience/definition/first-call-resolution-FCR">FCR</a>), average handle time (AHT), average speed to answer, repeat call rate, calls answered per hour and agent utilization rate. If a contact center, for example, strives for FCR but also expects low AHT, it might be disappointed. The goal of FCR is to resolve customer issues with one phone call, eliminating the need for repeat calls and increasing customer satisfaction. But AHT might be longer as agents work to address the problem.</p> <h3>3. Provide feedback to agents on 100% of monitored calls</h3> <p>For calls that businesses monitor via analytics, a scorecard, which measures customer service and agent performance, should be sufficient. However, companies should <a href="https://www.techtarget.com/searchcustomerexperience/tip/How-to-manage-remote-call-center-agents">provide agents with timely feedback and coaching</a> on monitored calls instead of waiting for a monthly review. It's also important for companies to provide agents direct feedback from customers. Companies need to offer agents feedback and coaching in areas of strength and opportunity. Some contact center platforms now use AI-driven coaching tools that automatically identify performance trends and recommend targeted training opportunities for agents.</p> <h3>4. Enable agents to listen to and score their own phone calls</h3> <p>In many cases, agents are the toughest critics of their own work. They should have the opportunity to hear how they sound and interact with customers.</p> <h3>5. Include side-by-side monitoring</h3> <p>Side-by-side monitoring enables analysts and supervisors to interact with agents and ask questions immediately following a phone call. Contact center management can then gather additional valuable insight into specific actions during the customer interaction, including any gaps in the tools agents use.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineImages/searchcrm_callcenter.jpg"> <img data-src="https://www.techtarget.com/rms/onlineImages/searchcrm_callcenter_mobile.jpg" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineImages/searchcrm_callcenter_mobile.jpg 960w,https://www.techtarget.com/rms/onlineImages/searchcrm_callcenter.jpg 1280w" alt="Two contact center agents wearing headsets review information on a computer screen while other agents work at nearby stations."> <figcaption> <i class="icon pictures" data-icon="z"></i>Supervisors and analysts often review agent interactions together during contact center monitoring to evaluate performance and identify coaching opportunities. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <h3>6. Use a different quality form for each customer service channel</h3> <p>Contact centers interact with customers across multiple channels, including phone, email, mobile apps, chat and social media. It's necessary to <a href="https://www.techtarget.com/searchcustomerexperience/tip/Best-practices-for-contact-center-quality-assurance">create different QA forms for each channel</a> to gather appropriate insights. On a QA monitoring form for phone calls, for example, one question might be about an agent's active listening skills. While that question is appropriate for a phone call, it might not provide any value for an email interaction.</p> <h3>7. Save examples of excellent customer interactions</h3> <p>Contact center managers monitoring agent performance inevitably come across some examples of excellent service and support that should be saved for later review and shared during training sessions. Contact centers can use these gold-standard examples as training tools for new agents and <a href="https://www.techtarget.com/searchcustomerexperience/answer/Nine-skills-every-call-center-agent-job-requires">agents who need to brush up on their skills</a> by highlighting the language and techniques that helped create outstanding CX.</p> </section> <section class="section main-article-chapter" data-menu-title="Technologies to support advanced contact center monitoring"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Technologies to support advanced contact center monitoring</h2> <p>A basic contact center monitoring program requires a technical foundation of quality monitoring software, which is included in many <a href="https://www.techtarget.com/searchcustomerexperience/definition/contact-center-as-a-service-CCaS">contact center-as-a-service</a> platforms and provided as a standalone tool by many vendors. This technology enables a team to listen to a sample of recorded phone calls and score each one using an electronic form.</p> <p>The first step in enhancing a monitoring program is to add the capability of capturing contact center agents' computer screens when recording a call. Screen captures enable analysts to do the following:</p> <ul type="disc" class="default-list"> <li>Observe how agents interact with desktop tools.</li> <li>Identify areas where agents can improve a process or transaction.</li> <li>Determine how businesses can improve desktop systems and tools to streamline processes and improve CX.</li> </ul> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineImages/crm-call_center_technologies.jpg"> <img data-src="https://www.techtarget.com/rms/onlineImages/crm-call_center_technologies_half_column_mobile.jpg" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineImages/crm-call_center_technologies_half_column_mobile.jpg 960w,https://www.techtarget.com/rms/onlineImages/crm-call_center_technologies.jpg 1280w" alt="List of technologies aiding contact center monitoring" height="292" width="279"> <figcaption> <i class="icon pictures" data-icon="z"></i>Speech analytics software facilitates contact center monitoring. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>The next step is to use <a href="https://www.techtarget.com/searchcustomerexperience/definition/speech-analytics">speech analytics</a> software to increase the number of calls monitored without requiring more staff to perform the function. Speech analytics can help increase the volume of quality monitors, especially at the agent level, and automate the call scoring process. With an increased number of monitors, patterns showing where an agent may be struggling with a specific type of inquiry can be more easily identified.&nbsp;</p> <p>Speech analytics provides several benefits beyond the ability to monitor a higher volume of calls. It can be used to identify the root cause of phone calls, which is more effective than analyzing disposition codes entered by an agent. Businesses can run a query, for example, that provides 100 calls in which customers have similar issues with a product. Analysts can listen to those calls, identify the root cause of a problem with a product or service, and resolve it.&nbsp;</p> <p>Speech analytics can also analyze phone calls for specific words, phrases, patterns and tones and provide reports. A word cloud, for example, is a collection of words depicting the frequency they appear in calls so companies can better identify customer expectations and sentiment communicated during calls. In more advanced real-time speech analytics, AI analytical capabilities are used in real time to identify calls in which the agent or <a target="_blank" href="https://www.dialora.ai/blog/ai-voice-frustration-detection-call-centers" rel="noopener">customer is becoming frustrated</a> and notify a supervisor to assist in handling the call.</p> <p>Many modern contact center platforms also incorporate real-time agent assist tools that analyze conversations during live calls and recommend next best actions.</p> <p>As contact center technology evolves, monitoring programs are becoming more data-driven and automated. Organizations that combine traditional monitoring practices with modern analytics tools can gain deeper insight into customer behavior and service gaps.</p> <p><b>Editor's note:</b> <i>This article was updated to reflect the latest developments in contact center monitoring tools, techniques and practices.</i></p> <p><i>Scott Sachs is president and founder of SJS Solutions, a consultancy that specializes in contact center strategy assessments and technology selection.</i></p> </section> A well-designed monitoring program identifies customer pain points and gathers valuable intelligence that can improve agent performance and CX, as well as products and services. https://cdn.ttgtmedia.com/rms/onlineimages/customer_service03.jpg https://www.techtarget.com/searchcustomerexperience/tip/Best-practices-for-call-center-monitoring Thu, 09 Apr 2026 10:00:00 GMT <p>Executive leaders should treat compliance as an integral part of organizational strategic planning rather than the cost of doing business.</p> <p>Organizations can face major penalties if they don't comply with laws and regulations that protect customer data, like GDPR and HIPAA. Additionally, customers can lose confidence in an organization if their personal information is not protected properly.</p> <p>A significant number of <a href="https://www.techtarget.com/searchcustomerexperience/tip/How-to-improve-the-contact-center-experience-for-customers">customer interactions occur in the contact center</a>. Therefore, contact center leaders need to comply&nbsp;with regulatory requirements and smart business practices to protect customers' rights. A combination of following legislative rules and thoughtful internal practices -- such as call monitoring -- can protect sensitive customer data while improving the customer experience.</p> <p>Establishing and following a contact center compliance checklist provides a strong foundation of good practices that lead to successful compliance.</p> <section class="section main-article-chapter" data-menu-title="What is contact center compliance and why is it important?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is contact center compliance and why is it important?</h2> <p>Contact center compliance is critical. A failure, such as a data breach, can have significant negative effects on a customer's life and devastate an organization's brand image and reputation. Customers don't want to buy services from organizations that can't&nbsp;protect their personal information from bad actors. And, if there's a security incident, organizations don't want to pay fines and penalties to regulatory agencies.</p> <p>Compliance requires participation from every individual in an organization. Contact center managers shouldn't assume that documented processes always work or that agents always follow proper procedures. <a href="https://www.techtarget.com/searchcustomerexperience/tip/Best-practices-for-call-center-monitoring">Ongoing monitoring and reporting must be in place</a> to ensure things are working properly. Additionally, all employees must keep their eyes and ears open. Controls must be in place, and if something does not seem right, they must raise the issue with the appropriate individual.</p> <p>Before the COVID-19 pandemic, most contact centers were on-premises, which, in many ways, made compliance easier to implement and monitor. For example, employees had to swipe their key cards to enter the contact center. Compliance became more of a challenge when contact centers began to operate remotely, so checklists can help contact center managers follow proper guidelines.</p> </section> <section class="section main-article-chapter" data-menu-title="Contact center compliance checklist"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Contact center compliance checklist</h2> <p>Organizations can't achieve compliance with a single tool or process. Compliance requires a multifaceted approach that integrates technology, processes and procedures.</p> <p>The following contact center compliance checklist can serve as a starting point for contact center managers as they seek to comply with internal and external requirements.</p> <h3>1. Secure the network</h3> <p>Organizations should use network&nbsp;<a href="https://www.techtarget.com/searchsecurity/definition/access-control">access control</a>&nbsp;to limit who can physically and logically access system hardware and software. Physical security protects the physical components of a network, such as devices, modems or routers, from physical harm. Logical security uses passwords and system permissions to protect a network's software and data from unauthorized individuals.</p> <h3>2. Lock down workstations</h3> <p>For remote workers, organizations must ensure workstation equipment adheres to pre-defined specifications or that the organization provides the proper tools.</p> <p>Physical workstation audits enable an organization to inspect both on-site and remote employees' work environments and ensure they support basic controls and meet compliance requirements. As physical visits to employees' remote workstations aren't always feasible, supervisors can use video conferencing to perform high-level audits. Beware: A video conferencing audit is limited in its scope and timing.</p> <h3>3. Authenticate customers</h3> <p>Customer authentication is a process where individuals prove they are who they claim to be. In some cases, single-factor authentication, where customers provide a single piece of information to confirm their identity, can suffice. However, many organizations have adopted&nbsp;<a href="https://www.techtarget.com/searchsecurity/definition/multifactor-authentication-MFA">multifactor authentication</a>, which asks customers to provide distinct pieces of information -- such as a password and a code sent to a mobile device -- to confirm their identity. Additionally, some companies are using voice and&nbsp;<a href="https://www.techtarget.com/searchenterpriseai/definition/facial-recognition">facial recognition</a>&nbsp;technologies to authenticate customers.</p> <h3>4. Record customer conversations</h3> <p>Call recording lets organizations&nbsp;review telephone conversations&nbsp;between customers and agents. Managers can review recordings through a QA program and AI tools to determine if agents fulfilled external requirements, such as appropriate disclosures and authentication processes. Managers can also review recordings to determine if an agent fulfilled internal requirements, such as providing a customer with accurate information or following the correct internal procedures.</p> <h3>5. Provide mandatory disclosures</h3> <p>Contact center agents must provide mandatory disclosures, which are legal statements to explain specific processes, rules and options to callers. For example, if a contact center in the U.S. wants to record a customer call, it must be disclosed to the caller, and consent must be provided, which is passive consent in most cases. Regulations require mandatory disclosures when agents record customer calls, perform collection functions or make financial transactions.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/call_center_compliance_checklist-f.png 1280w" alt="Contact center compliance checklist image" height="266" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>This compliance checklist can help contact centers adhere to important standards and regulations. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <h3>6. Adhere to local privacy legislation</h3> <p>Organizations must adhere to various global and local legislation on customer privacy, depending on the geographic reach of the business. Due to legislation, organizations can't manage all customer information in the same way.&nbsp;<a href="https://www.techtarget.com/searchsecurity/tip/State-of-data-privacy-laws">Data privacy laws vary by country and region</a>, so organizations must know where each customer resides before they transmit, process and store customer information.</p> <p>Examples of location-based privacy legislation include the following:</p> <ul type="disc" class="default-list"> <li><b>General Data Protection Regulation.</b>&nbsp;GDPR provides guidance on how organizations can collect and process personally identifiable information (PII) for individuals who live in the European Union.</li> <li><b>California Consumer Privacy Act.&nbsp;</b>CCPA provides guidance on how organizations can collect and process personal and household information for individuals who live in California.</li> </ul> <p>Because privacy legislation is always evolving, organizations must be proactive to ensure they're up to date on laws affecting their contact center operations.</p> <h3>7. Adhere to the Telephone Consumer Protection Act</h3> <p>Organizations in the U.S. must adhere to the Telephone Consumer Protection Act, which sets rules for how an organization can use outbound calls for solicitation.&nbsp;<a target="_blank" href="https://www.fcc.gov/sites/default/files/tcpa-rules.pdf" rel="noopener">TCPA</a>&nbsp;regulations state that telemarketing contact centers cannot use predictive dialers to contact a wireless phone without prior consent from the customer. It also ensures telemarketers adhere to the National Do Not Call Registry and special regulations, which may include restricted calling hours in a particular geographic location after a natural disaster event.</p> <h3>8. Manage sensitive information</h3> <p>To comply with standards, such as the&nbsp;<a href="https://www.techtarget.com/searchsecurity/definition/PCI-DSS-Payment-Card-Industry-Data-Security-Standard">Payment Card Industry Data Security Standard</a>&nbsp;and HIPAA, organizations must&nbsp;protect sensitive customer data&nbsp;at rest and in motion. Sensitive information can include PII, credit card numbers or protected health information. To protect sensitive information, organizations should encrypt all data, minimize the amount of stored data and use automation, such as&nbsp;<a href="https://www.techtarget.com/searchcustomerexperience/definition/Interactive-Voice-Response-IVR">interactive voice response</a>, to perform sensitive transactions.</p> <h3>9. Provide ongoing training</h3> <p><a href="https://www.techtarget.com/searchcustomerexperience/tip/Best-practices-for-call-center-agent-training-programs">Organizations should provide annual training</a>&nbsp;on proper compliance procedures and guidelines to all employees. All employees should be up to date on specific compliance rules and understand how they can protect their organization and its customers.</p> <h3>10. Promote self-service</h3> <p>Organizations should maximize <a href="https://www.techtarget.com/whatis/definition/customer-self-service-CSS">customer self-service</a> procedures through secured portals to limit the sharing of information with other individuals and reduce security risks.</p> <h3>11. Test, monitor and act on a continuous basis</h3> <p>Organizations can and should implement all the items on this checklist. However, business leaders shouldn't assume everything is working properly or that bad actors have not found ways to bypass established controls. Organizations should continually test and monitor the various compliance controls, whether through automated processes, such as reporting unauthorized attempts to access customer data, or human processes, like placing test calls.</p> <p>When something doesn't seem right, organizations should analyze the issue and take action to ensure the controls in place are performing as expected.</p> <p>A contact center compliance checklist can help organizations&nbsp;<a href="https://www.techtarget.com/searchcio/feature/9-common-risk-management-failures-and-how-to-avoid-them">avoid compliance failures</a>. Contact center managers can use this checklist to evaluate their organization's current compliance protocols and ensure their teams follow proper guidelines.</p> </section> <section class="section main-article-chapter" data-menu-title="Common contact center compliance issues and malpractices"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Common contact center compliance issues and malpractices</h2> <p>If strong compliance controls and practices are not in place, the following negative events can occur:</p> <ul type="disc" class="default-list"> <li>Calling customers who requested not to be called, which violates&nbsp;<a href="https://www.techtarget.com/searchcustomerexperience/definition/outbound-call">outbound calling</a>&nbsp;restrictions.</li> <li>Allowing PII to be stolen due to incorrectly accessing customer information.</li> <li>Improperly recording customer conversations by not adhering to call recording and consent rules.</li> <li>Providing incomplete information to customers by not adhering to scripting requirements.</li> </ul> <p>If these events occur, an organization can be liable for financial penalties, along with other legal consequences. Just as important is the potential for negative customer perception, which can lead to degraded customer loyalty and customer defections.</p> <p>Leadership in all areas of an organization must keep in mind that a compliance checklist only provides a template of best practices. It must be more than a written document. The organization must translate the document into reality and embed it into the corporate culture by focusing on the following actions:</p> <ul class="default-list"> <li>Invest in technology to support key items on the checklist.</li> <li>Promote accountability.</li> <li>Reward adherence to policies and address any gaps that arise.</li> </ul> <p>A contact center compliance checklist might not stop all unauthorized activities, but it's a solid start to implementing a strategy that adheres to legal, organizational and customer requirements.</p> <p><i>Scott Sachs is president and founder of SJS Solutions, a consultancy that specializes in contact center strategy assessments and technology selection.</i></p> </section> A contact center compliance checklist can serve as a starting point for contact center managers as they seek to comply with internal and external regulations. https://cdn.ttgtmedia.com/rms/onlineimages/check_g1268128622.jpg https://www.techtarget.com/searchcustomerexperience/tip/Call-center-compliance-checklist-for-hybrid-workforces Thu, 02 Apr 2026 09:00:00 GMT <p>Increasingly sophisticated adversaries are putting IT on the defensive. A cohesive approach to network security is more critical than ever.</p> <p>Threat actors have been quick to adopt cutting-edge technologies, among them <a href="https://www.techtarget.com/searchenterpriseai/tip/How-to-manage-generative-AI-security-risks-in-the-enterprise">AI and automation</a>, to make their attacks more potent. At the same time, cybercriminals are more aggressive, putting more pressure on the network security practitioners standing on the front lines in defense of enterprise assets. To combat these threats, organizations must craft a comprehensive, scalable network security management strategy that uses best practices to protect their network from end to end.</p> <p>Threat actors profit by using a combination of tactics, including phishing, ransomware and AI-generated deepfakes, to breach organizations.&nbsp;IBM's 2025 <a href="https://www.ibm.com/downloads/documents/us-en/131cf87b20b31c91">report</a> on the Cost of a Data Breach pegged the average cost of a data breach in the U.S. at upward of $10 million -- driven by more stringent regulatory fines and a surge in detection expenses.&nbsp;</p> <p>In this volatile climate, security practitioners must arm themselves with tools that protect the organization before attacks occur, as well as platforms that mitigate incidents as quickly as possible. Resilience is the ultimate objective.</p> <section class="section main-article-chapter" data-menu-title="What is effective network security management?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is effective network security management?</h2> <p>A solid network defense underpins every productive and healthy enterprise network. The most effective strategies map the <a href="https://www.techtarget.com/searchsecurity/feature/How-to-create-a-data-security-policy-with-template">right security policies</a>, tools, processes and practices to the organization's operational objectives.</p> <p>Enterprises must also heed government regulations and corporate mandates. Security practitioners are tasked with ensuring data integrity, security and the availability of their infrastructures. Yet, attaining complete protection isn't achievable. No enterprise can lock down its environment 100% without sacrificing productivity.</p> <p>To that end, effective enterprise security taps into essential technologies, such as the following:</p> <ul type="disc" class="default-list"> <li>AI-driven endpoint security platforms.</li> <li>Extended detection and response.</li> <li>Platformization.</li> <li>Zero-trust and access management.</li> <li>Firewalls.</li> <li>Multifactor authentication (MFA).</li> <li>Identity and access management.</li> <li>Network and <a href="https://www.techtarget.com/searchnetworking/definition/network-analytics">traffic monitoring and analysis</a>.</li> <li>Security information and event management.</li> <li>Identity and access management.</li> <li>Vulnerability and testing.</li> <li>Unified threat management.</li> <li>DoS mitigation and incident response services.</li> <li>Encryption.</li> <li>Data loss prevention.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Network security management challenges"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Network security management challenges</h2> <p>Securing the infrastructure is a top priority, but security teams tasked with protecting network assets face some high hurdles, such as the following:</p> <ul class="default-list"> <li>Virtual and highly distributed enterprise assets.</li> <li>Network security controls that may impede infrastructure performance and hamper the end-user experience.</li> <li>The need to integrate security data from disparate sources to protect hybrid environments and other network designs.</li> </ul> <ul type="disc" class="default-list"></ul> <p>Network security tools are continually improving, but certain issues still force network security engineers to scramble as they try to stay ahead of threats.&nbsp;It's still a challenge to obtain an accurate end-to-end perspective of network activity from multiple sources, particularly in&nbsp;<a href="https://www.techtarget.com/searchitoperations/tip/Navigate-hybrid-cloud-observability-with-3-techniques">hybrid cloud environments</a>.&nbsp;Even in products that supposedly have close correlation, true integration is often missing.</p> <p>Improvements in threat identification have helped. The 2025 IBM Data Breach Report found that a combination of AI and automation helped security practitioners correctly identify breaches 80 days faster than the prior year.</p> <p>Despite these gains, the reality for most organizations is that it's a question of when -- and not if -- they will be breached. That's one reason why so many have <a href="https://www.techtarget.com/searchsecurity/feature/How-to-implement-zero-trust-security-from-people-who-did-it">embraced zero-trust architectures</a> designed to ensure that only authenticated and authorized users gain network access.</p> <p>Zero-trust employs many protective controls, including granular authentication, which considers the following factors before allowing any entity or person access to the network:</p> <ul type="disc" class="default-list"> <li>User identity.</li> <li>Device type.</li> <li>Activity.</li> <li>Query.</li> <li>Location.</li> </ul> <p>Zero-trust also applies ongoing authentication and tracking to monitor users and devices. This ensures they have not been compromised.</p> <p>Another security tactic gaining favor is segmentation, which limits network access and prevents lateral movement. Organizations can also implement&nbsp;<a href="https://www.techtarget.com/searchsecurity/answer/Compare-zero-trust-vs-the-principle-of-least-privilege">least-privilege access</a>, which applies MFA and granular micro-segmentation to further control access to enterprise resources.</p> <p>Network observability, in which IT administrators gain insights beyond traditional network monitoring by seeing and investigating activity in real time, is <a href="https://www.techtarget.com/searchitoperations/tip/Observabilitys-role-in-mitigating-IT-security-risks">another key advance</a> helping organizations mount stronger defenses. Using both security and network performance intelligence, observability helps security engineers better discern suspicious activity, optimize service levels and mitigate incidents.</p> <p>Finally, threat detection has become a more powerful defensive tool, thanks to incremental advances in machine learning. ML establishes a baseline of network behavior, observing when activity deviates. ML-based threat detection can distinguish between harmless anomalies and real threats.</p> </section> <section class="section main-article-chapter" data-menu-title="Effective network security management best practices"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Effective network security management best practices</h2> <p>The best network security technology is only as effective as the policies and practices that implement controls.&nbsp;IT teams and end users must be aware of the protections in place and how to use them effectively.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Network security tools are continually improving, but certain issues still force network security engineers to scramble as they try to stay ahead of threats. </figure> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>This starts with proper training. All employees and contractors should understand&nbsp;<a href="https://www.techtarget.com/searchsecurity/tip/How-to-write-an-information-security-policy-plus-templates">corporate IT security policies</a>&nbsp;and how to use available tools. Policy development and continuous review is a crucial foundational element.&nbsp;Ongoing end-user training and education -- not just an annual training session or quiz -- is a must.</p> <p>Other fundamental best practices include the following:</p> <ul type="disc" class="default-list"> <li>Deploy multilayered security with support and countermeasures. This limits lateral access and protects the most critical resources.</li> <li>Have an effective network monitoring service in place.</li> <li>Regularly update both software and hardware. Change factory hardware settings when deploying new equipment.</li> <li>Automate software patching but manually intervene when necessary.</li> <li>Consider employing automation for workflow optimization and some elements of remediation.</li> <li>Look into AI-assisted response.</li> <li>Perform vulnerability assessments on a consistent basis. Conduct interim testing between audits.</li> <li>Apply MFA and other access controls.</li> <li>Implement network segregation.</li> <li>Develop and deploy a zero-trust architecture that relies on continuous verification.</li> <li><a href="https://www.techtarget.com/searchsecurity/feature/5-critical-steps-to-creating-an-effective-incident-response-plan">Define an incident response plan</a>&nbsp;that applies automation when possible.</li> <li>Take steps to prevent insider theft or data loss.</li> <li>Recognize what baseline network activity looks like.</li> <li>Proactively test systems to uncover vulnerabilities and poor configurations.</li> </ul> <p>Effective network security management best practices start and end with the human element.&nbsp;The most proactive organizations understand this concept and ensure the right practices and processes are in place.</p> <p><em>Amy Larsen DeCarlo has covered the IT industry for more than 30 years, as a journalist, editor and analyst. As a principal analyst at GlobalData, she covers managed security and cloud services.</em></p> </section> Threat actors are using increasingly sophisticated tools to make their attacks more costly. It's time for organizations to craft a comprehensive security management strategy. https://cdn.ttgtmedia.com/visuals/searchSoftwareQuality/security_testing/softwarequality_article_015.jpg https://www.techtarget.com/searchnetworking/answer/How-are-network-management-and-security-converging Fri, 27 Mar 2026 11:15:00 GMT <p>Nvidia GTC 2026 was bigger and broader than ever, reflecting the continued, relentless growth of demand and rapid pace of change and innovation in the AI sector. Nvidia has placed itself at the center of it all, as the key supplier of a growing number and variety of xPUs and the related systems that underpin the most advanced AI factories.</p> <p>Equally important is the size of the ecosystem that Nvidia has cultivated across the full AI stack, as reflected by the 400+ exhibitors on the exposition floor.&nbsp;</p> <section class="section main-article-chapter" data-menu-title="Nvidia GTC highlights"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Nvidia GTC highlights</h2> <p>Nvidia CEO Jensen Huang highlighted several important developments from the Nvidia mainstage keynote, including the following:</p> <ul class="default-list"> <li><b>NemoClaw overview.</b> NemoClaw, a safe environment for OpenClaw built on OpenShell, offers enterprises a path to safely use the approach to <a href="https://www.techtarget.com/searchenterpriseai/tip/How-to-build-your-first-agentic-AI-system">building and running AI agents</a>.</li> <li><b>Token production efficiency. </b>Incorporating Nvidia Groq 3 LPX into the AI factory substantially extends token production velocity at the high end of the range, overcoming the inherent limitations of NVL72.</li> <li><b>Storage and performance optimization. </b>STX, the new storage reference architecture enabled by BlueField-4, reinforced data storage as an essential part of the mix that must be optimized -- along with compute and networking -- to continue <a href="https://www.techtarget.com/searchstorage/opinion/Strap-in-for-the-next-wave-of-AI">driving capacity up and incremental cost down</a>.</li> <li><b>Revenue growth. </b>Nvidia predicts<b> </b>it could generate $1T in visible revenue potential by the end of 2027. Growth will partly be due to more AI Gigafactory builds, but increasingly it will be driven by the rapid expansion in infrastructure demand for distributed inferencing. Nvidia believes the long-anticipated pivot point has been reached, and training will continue to drive revenue; AI inferencing will drive them faster from this point forward.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Tokenomics as a driver for AI business cases and ROI"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Tokenomics as a driver for AI business cases and ROI</h2> <p>The AI sector is focused on <a target="_blank" href="https://medium.com/@nconleth/understanding-tokenomics-1ae2e2c2fb05" rel="noopener"><i>tokenomics</i></a>, a term with roots in blockchain and cryptocurrency but applicable to AI as well. In the world of AI, tokens are the unit of production. The capacity and performance of AI systems are measured in terms of time to first token (TTFT) and tokens per second.</p> <p>Contextual data management is measured by token retention rate. There are even efforts underway to track value per token to understand and balance quality. And, increasingly, AI services are being metered and licensed in tokens.&nbsp;</p> <p>According to Huang, every chief experience officer should be thinking about how much token capacity to grant each employee, and employees should be measured by how well they use their token budgets. This concept represents a means by which <a href="https://www.techtarget.com/searchenterpriseai/feature/10-AI-business-use-cases-that-produce-measurable-ROI">AI technology usage can be measured, managed and translated into business value</a>.</p> <p>Proper management and review of tokens, as with any other scarce resource, separate good use from bad. The bottom line is that everyone needs to start thinking in terms of tokenomics for business cases, KPIs and ROI across the ecosystem, from both supplier and user perspectives.</p> </section> <section class="section main-article-chapter" data-menu-title="Ecosystem highlights"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Ecosystem highlights</h2> <p>The conference featured many press releases and announcements from vendors in Nvidia's ecosystem. Findings from discussions related to enterprise networking infrastructure include the following:</p> <h3>Akamai</h3> <p>Akamai announced support for the Nvidia AI Grid reference architecture through its Akamai Inference Cloud, based on the deployment of thousands of Nvidia Blackwell GPUs and BlueField DPUs across its 4,400 global edge locations.&nbsp;</p> <h3>Arrcus</h3> <p>Arrcus promoted its recently announced Arrcus Inference Network Fabric, a purpose-built networking approach for optimizing the delivery of inferencing workloads across distributed environments, now integrated with Nvidia Dynamo, Nvidia BlueField-3 DPUs and Nvidia Spectrum-X Ethernet.</p> <h3>Cisco</h3> <p>Cisco announced several new products related to its Secure AI Factory initiative to simplify network design, deployment and operations. Products include the following:</p> <ul class="default-list"> <li>UCS and Unified Edge support for Nvidia RTX Pro Blackwell GPUs.</li> <li>Reference architecture support for AI Grid.</li> <li>A new N9100 switch powered by Nvidia Spectrum-6 silicon.</li> <li>Integrated support of that same N9100 switch series within Cisco Nexus Hyperfabric.</li> </ul> <p>On the security side, Cisco announced the extension of its Hybrid Mesh Firewall for policy enforcement on Nvidia BlueField DPUs connected to Cisco Nexus ONE fabrics, as well as new purpose-built guardrails for AI agents within Cisco AI Defense, including support for Nvidia OpenShell for safer <a href="https://www.techtarget.com/searchcio/feature/OpenClaw-and-Moltbook-explained-The-latest-AI-agent-craze">OpenClaw deployments</a>.</p> <h3>Cloudflare</h3> <p>Cloudflare was highlighting its offerings to directly support the development of stateful serverless AI agents, using its global Connectivity Cloud network reach and Agent Developer Platform, which includes a deep model repository, durable objects, and egress-free R2 storage for optimizing both cost and performance.</p> <h3>Dell</h3> <p>Dell continued to advance its AI Factory, with over 4,000 customers now in deployment and early adopters seeing ROI increases of nearly three times in the first year. The product uses Dell's ability to bring together compute, storage and networking, including PowerEdge servers that use both Nvidia HGX Rubin NVL8 and Vera Rubin NVL72 GPUs for high-end training, as well as RTX Pro 4500 Blackwell GPUs and the new Vera CPU for enterprise workloads in the data center.&nbsp;</p> <h3>Equinix</h3> <p>Equinix highlighted its recently launched Distributed AI Hub, part of the Equinix Fabric Intelligence initiative, and its first direct partnership with Palo Alto Networks to simplify and secure agentic AI deployments.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> The bottom line is that everyone needs to start thinking in terms of tokenomics for business cases, KPIs and ROI across the ecosystem. </figure> <i class="icon" data-icon="z"></i> </div> </blockquote> <h3>F5</h3> <p>F5 has turned its focus onto improving tokenomics outcomes -- such as TTFT, token throughput and cost per token -- by integrating its BIG-IP Next for Kubernetes with Nvidia BlueField-3 DPUs. Given that most AI applications are built using cloud-native architectures, this approach acts as a control plane that optimizes results in a manner closely aligned with where workloads reside and execute.</p> <h3>HPE</h3> <p>HPE announced new Nvidia-related capabilities across compute, networking and software realms, set to be available in 2027. Examples include integration of Nvidia Vera CPU blades and Nvidia Quantum-X800 InfiniBand networking into the HPE Cray Supercomputing GX5000 platform.</p> <p>HPE also updated its turnkey HPE Private Cloud AI offerings to expand network capacity and introduced ProLiant servers with Nvidia RTX Pro 6000 Blackwell GPU support. Additionally, the company also integrated Nvidia AI-Q and Omniverse into its design blueprints, offering self-contained options for sovereign AI factory deployments.</p> <h3>IBM</h3> <p>IBM relayed results of using Nvidia cuDF to accelerate its watsonx.data's SQL engine Presto to dramatically accelerate terabyte-scale data refresh cycles for a global foods supplier. This yields a remarkable 30x improvement in price-to-performance over prior processes. IBM also promoted its certification of IBM Storage Scale 6000 with Nvidia DGX platforms to address ongoing challenges with AI data management and storage.</p> <h3>Lenovo</h3> <p>Lenovo has been working to deliver AI infrastructure in multiple forms, from Gigafactory architectures to laptops and workstations powered by Nvidia RTX Pro Blackwell GPUs for local inferencing. It also launched an industry-vertical platform focused on improving fan experience, revenue growth, business performance and operational efficiency for global sports organizations.</p> <h3>Nokia</h3> <p>Nokia announced a slew of new optical components and amplifiers for use within and between data centers, coming available later this year and into 2027.&nbsp;</p> <h3>Nutanix</h3> <p>Nutanix focused on addressing challenges with <a href="https://www.techtarget.com/searchenterpriseai/tip/Optimizing-hybrid-cloud-architecture-for-AI-workloads">deploying AI workloads in mixed environments</a>, where both cloud-native and virtual compute components will be present. The company also announced its Nutanix Agentic AI platform, a full software stack designed to help accelerate agent development, security and operational deployment.</p> <h3>WEKA</h3> <p>WEKA announced the general availability of the NeuralMesh AI Data Platform, which aims to make scaling high-performance data pipelines seamless as AI workloads move from PoC to production. WEKA also announced the Augmented Memory Grid, which provides accelerated data pipelining for existing Nvidia-based AI factories, similar to how the new Nvidia STX architecture will work with future reference architectures.</p> </section> <section class="section main-article-chapter" data-menu-title="The AI wave continues"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The AI wave continues</h2> <p>When attending an event built around an individual technology vendor, it's usually easy to see through even the most carefully constructed frameworks of stories and promises. But when those stories come from a chorus of surrounding voices, it's much more convincing.</p> <p>Getting the market timing right is one thing, and it has a lot to do with Nvidia's success, following years of tireless devotion to PC GPU cards. But one must also credit Nvidia's vision and execution in building and nurturing an ecosystem that can sustain and compound opportunities, sharing both the risks and the rewards.</p> <p>Lingering questions about whether the massive AI buildout is a bubble or a wave remain. <a href="https://www.techtarget.com/whatis/feature/Explaining-an-AI-bubble-burst-and-what-it-could-mean">Bubbles can grow very large but fail to deliver and pop</a> when overhyped based on hype rather than measurable value. Waves also start large but carry on due to a successful translation into outcomes.</p> <p>AI is a big technology wave, and it's not showing any signs of slowing. In fact, quite the opposite is happening -- it's still building. The positive value outcomes are already being documented, and we now stand on the threshold of the era of AI inferencing, where the real and measurable values start rolling in, in ever-increasing volume.&nbsp;&nbsp;</p> <p><em>Jim Frey covers networking as principal analyst at Omdia.</em></p> <p>&nbsp;<em>Omdia is a division of&nbsp;Informa TechTarget.&nbsp;Its analysts have business relationships with technology vendors.</em></p> </section> Nvidia GTC 2026 underscored surging AI infrastructure demand and ecosystem growth. Tokenomics and inferencing are emerging as key drivers of future value. https://cdn.ttgtmedia.com/rms/onlineimages/security_a244600171.jpg https://www.techtarget.com/searchnetworking/opinion/Infrastructure-highlights-from-Nvidia-GTC-2026 Thu, 26 Mar 2026 14:00:00 GMT 60 webmaster@techtarget.com