Alteryx on Wednesday unveiled VURA, a new governance experience in its Alteryx One platform that provides users with a control layer to oversee their AI tools and validate AI outputs.
VURA — which stands for visible, understandable, repeatable and auditable — is designed to help Alteryx customers better understand how AI tools arrive at their responses and make sure those responses can be trusted.
In May, a survey commissioned by Alteryx report found that 43% of respondents report that interpreting and explaining AI-generated insights remains one of the main barriers still preventing them from putting agents and other AI tools into production.
The visible aspect of VURA addresses uncertainty by allowing users to trace AI outputs to their data sources to see how AI tools reason and reach conclusions; understandable provides asset descriptions, metadata, certifications and quality information to demonstrate the underlying logic used by AI; repeatable ensures consistency by making sure that AI tools only access governed, reliable business logic; and auditable makes visible formal audit trails and documentation so human users can ensure proper governance and compliance before approving AI-generated suggestions.
David Menninger, an analyst at ISG, noted that in his firm's most recent study on AI, nearly three-quarters of all respondents reported that security, governance and privacy concerns limit the value they're getting from their investments in AI. In addition, the survey found that governance alone accounts for 11% of organizations' spending on AI.
In addition, while ISG's research reflects ongoing concerns around governance in AI, Deloitte in July 2025 found that trust is the biggest barrier to agentic AI adoption in the finance and accounting industries, with less than 5% of respondents reporting that they trust autonomous agents to always make decisions. Similarly, McKinsey reported in March 2026 that enterprises are making progress in trust maturity, but gaps in strategy, governance and risk management are ongoing problems.
"The additional visibility and documentation that [new] VURA capabilities provide will help address some of those concerns," Menninger said.
In addition, while AI governance capabilities that address visibility and control are not unique, Alteryx's positioning as a preparation specialist that sits between sources of data and consumption of data somewhat distinguishes what the vendor is providing from what competitors such as Informatica and Databricks offer, he continued.
"It plays the role of middleman and has visibility into the entire data estate and the flows of data within the data estate," Menninger said. "As you apply AI to those flows, it can provide a pretty thorough explanation of the data, and it can create controls over the access and use of the data."
Building believability in AI
A combination of what Alteryx was hearing from its own customers plus observing what is happening broadly as enterprises try to scale AI drove Alteryx to add the new VURA governance capabilities, according to Ben Canning, the vendor's chief product officer.
"The issue holding back AI adoption isn't capability, it's trust," he told TechTarget. "Our customers told us the bottleneck isn't getting an answer, it's defending one. … VURA lets the business certify, approve and govern its data and logic, then makes those trust signals available to the AI itself so the answer arrives with its receipts."
Following the launch of the new AI governance capabilities, Alteryx plans to continue making its platform a business logic layer for AI, Canning continued. Specific initiatives include enabling users to build a governed registry of their enterprise's business logic and capabilities that spot when users query the wrong data and steer them to a better source.
"Every CFO has been told to use AI. Almost none of them can show the board how the AI got the number. That’s the gap we’re closing," Canning said.
Beyond its four foundational pillars, VURA enables system administrators to do the following:
- Define software development lifecycle processes, versioning controls and certification criteria to govern the development of AI tools.
- Visualize and audit how datasets are transformed to see data lineage and ensure data quality.
- Integrate datasets with Google Cloud Knowledge Catalog, Atlan and Collibra so datasets are governed and traceable.
- Maintain semantic definitions and access controls through those same integrations with data catalogs.
- Set up identity permissions and other safety parameters that enable business users to generate visual workflow diagrams using natural language so domain experts can maintain the business logic that informs agents and other AI tools.
Donald Farmer, founder and principal of TreeHive Strategy, noted that governance is too often seen as restricting analysis rather than enabling creative and experimental insight. Alteryx's new VURA capabilities represent good governance for AI, according to Farmer.
"Well thought-out governance is an enabling function, not a constraint," he told TechTarget. "Alteryx is taking some interesting steps here to make this possible at scale."
In addition, by governing procedural business logic, Alteryx is providing somewhat different governance capabilities than its peers, Farmer continued.
"Making governance metrics available … has become standard now, but Alteryx is somewhat differentiated in what they are governing," he said.
Perhaps the most valuable features of Alteryx's VURA update are enabling users to define workflows and put controls in place, according to Farmer.
"I especially like the editing of visual workflow diagrams, enabled through natural language," he said. "That enables analysts to configure quite complex workflows, but IT will still enforce underlying identity management and guardrails."
However, domain-specific blueprints that help new customers use Alteryx One to govern AI would further improve VURA, Farmer continued.
"New customers struggle to build governed metric definitions and compliance pipelines from scratch," he said. "Shipping pre-built, 'auditor-reviewed' workflow packs for [specific] scenarios … could accelerate time-to-value."