The hidden cost of AI automation: Preserving organizational expertise

Enterprise software vendors are rapidly embedding AI agents and intelligent automation into ERP, HR, CRM, IT service management, collaboration and other enterprise platforms. While these capabilities promise greater efficiency by automating routine decisions and orchestrating workflows, they also raise an important governance question: How can organizations design AI-enabled enterprise workflows so that automation improves efficiency without weakening the human expertise needed to evaluate exceptions, correct errors and maintain operations?

Talent preservation: A new face of governance

From both corporate and legal perspectives, governance means that the business is accountable to its key stakeholders: employees, customers, shareholders and the broader community. The historical role of governance has been to reduce corporate risk. This risk was managed by maintaining the privacy of customer data, ensuring that data and other IT assets were secure, and working with users to set guardrails defining which systems and assets employees across functions are authorized to use.

However, with the introduction of AI, AI agents and greater business process automation, the enterprise risk management plane has broadened. How, for example, can enterprises maintain business resilience by ensuring there is no erosion of human skill sets and know-how as more AI and automation are introduced?

"We view this as an important topic that must be actively managed," said Christophe Theys, global head of AI, Data & Analytics for DHL Supply Chain. "There is a risk that people will become overly reliant on technology if AI is introduced without proper governance."

How organizations are adopting AI and preserving talent

AI delivers efficiencies to companies. It is helping employees work more effectively and improving company operations. As more AI is embedded in business processes, the largest gains from AI are still ahead.

Despite AI's success, the risk of workforce talent erosion due to increased use of AI and automation is not lost on enterprises, either. Here are three examples of companies that are advancing AI and ensuring their workforces remain skilled.

DHL

"We use AI across a broad range of operational and functional processes," said Theys. "Examples include warehouse operations, transportation management, customer service, staff functions and general employee productivity. We are increasingly deploying AI agents and embedded AI capabilities to automate repetitive tasks, support decision-making, improve service quality and help employees work more efficiently. Today, we have dozens of AI use cases globally, ranging from operational digital assistants and AI-powered customer interactions to AI agents supporting transportation and warehouse processes."

Theys said that most of the AI and AI agents employees are using are those embedded in enterprise systems. The company also develops some of its own.

"In all cases, our goal is a 'human plus AI' workforce where people focus more on exception handling, customer interaction, problem solving and continuous improvement," Theys said. "We invest in helping employees understand where AI can assist them, how to work effectively with AI tools and where human judgment remains essential."

The objective [of AI] is not to remove human knowledge, but to make it more accessible and more scalable.

DHL already sees measurable productivity improvements, faster response times, higher quality and better employee experiences. AI agents are automating many aspects of operational communication, including scheduling, status updates and some exception management, while keeping DHL employees in the loop.

"We see AI as an augmentation technology, not as a replacement for human expertise," Theys continued. "We deliberately keep humans involved in critical decisions, customer interactions, compliance-related activities and exception handling. We also focus on upskilling and knowledge sharing. In many cases, AI actually helps preserve organizational knowledge by making expertise easier to capture, document, search and reuse across the organization. The objective is not to remove human knowledge, but to make it more accessible and more scalable."

Kuehne+Nagel

"We are using AI across multiple areas of our business to improve productivity, decision-making and customer experiences," said Alireza Nemati, chief AI and innovation officer at logistics provider Kuehne+Nagel. "Our approach focuses on embedding AI into the way we work, rather than treating it as a standalone technology initiative. Today, AI supports activities such as pricing, booking, customer service, customs processing, workforce planning, workflow automation and employee productivity tools. We are also increasingly using AI-powered assistants and agents that help colleagues access information, automate tasks, make better decisions and spend more time on higher-value work."

Kuehne+Nagel integrates AI directly into the systems and workflows its employees use every day and develops its own AI capabilities while also using offerings from enterprise vendors.

"We view AI adoption as a business transformation, not simply as a technology rollout," said Nemati. "Capturing value from AI requires us to rethink how work is performed, redesign workflows with AI at their core and redefine roles, particularly in software development ... We are already seeing tangible benefits in several areas, including faster processing of operational tasks, higher employee productivity, quicker access to information, enhanced customer service and better decision support. AI is helping remove friction from workflows and allowing employees to spend more time on customer-facing and higher-value activities."

Nemati said that the goal of Kuehne+Nagel was to use AI to augment human expertise, not replace it. "While automation can take over repetitive and administrative tasks, logistics remains a complex business that depends on experience, judgment, customer understanding and the ability to manage exceptions. Human accountability, therefore, remains essential," said Nemati. Our approach is based on a 'human-in-the-loop' model, where AI supports decision-making while employees retain responsibility for outcomes."

Addison Group

Addison Group is an employment recruiting agency that uses AI to augment employee work, not replace it.

"We've implemented AI to automate manual processes and extend the capabilities of internal teams, particularly within shared services such as finance, HR, etc.," said Thomas Moran, CEO. "Our teams have integrated AI as a supplemental resource that operates alongside employees, with appropriate oversight to manage its use, review outputs and evaluate performance, ensuring work is complete, accurate and aligned with business requirements. In short, AI serves as an enhanced assistant to our business processes."

Addison has seen benefits as AI has streamlined business processes and improved efficiency, with the proviso that experienced professionals on staff have final oversight. "Done right, a company today may be able to double or triple the output of specific roles, with AI acting as a productivity and efficiency agent," said Moran.

Is there a risk that employee skills could erode over time as AI automates more work steps?

"It depends on how much automation is being asked for from AI," Moran said. "If the process for using AI is simply to enter a prompt and receive an answer without any further skill development, then yes, over time, erosion can happen. On the other hand, if AI is used as an enhancement that still allows room for human problem-solving and critical thinking, a company's workforce knowledge, in theory, can actually improve ... AI serves as an enhanced assistant to our business processes. When it comes to retraining, we focus on ensuring our team members understand AI's contributions to the process while highlighting where human expertise, critical thinking and connection remain essential for verification."

The bottom line

Most enterprises are learning as they go while adopting AI. A majority use AI embedded in the enterprise software they already run, while some augment these sources with their own AI. In all cases, the emphasis is on a human-in-the-loop strategy, with AI augmenting work processes to enhance human expertise rather than lose it.

Chris Miller, CEO of Point Quest Group, which provides special education staffing and services, also uses AI in the company's operations. Miller had this to say:

"[Employee] knowledge erosion becomes a risk only when leadership delegates judgment to AI instead of drudgery. If an organization leans on AI for strategic, diagnostic or relationship-driven decisions, workforce knowledge will erode. But if AI is confined to the structured and repetitive, such as pre-populating routine logs, flagging missing compliance fields and organizing raw data, it protects human capability … In our field, where highly trained specialists manage complex compliance and client care, administrative load and burnout are primary drivers of attrition. When senior practitioners spend their time on paperwork rather than the skills they were trained for, you lose institutional knowledge the moment they burn out and leave. Letting AI carry the administrative load gives those hours back to high-value human work. The rule of thumb is simple: AI handles the process load and the human owns every decision that requires nuance and judgment."

Summing it up

Enterprises understand that managing their human capital is an integral part of governance. Accordingly, as they adopt AI, they are also taking steps to preserve their workforce's skills and expertise. They are doing this in the following ways:

• Using AI to automate repetitive and mundane business processes that enable employees to do higher-level work.
• Adopting guidance that ensures that AI is not used without proper human oversight.
• Reskilling employees for new and transformative business processes that use AI.
• Making employee reskilling and education a continuous and not a one-off activity.