IBM study shows enterprise AI needs governance built in

Enterprise AI is running into a control problem, and IBM’s latest study makes that gap hard to ignore. As AI moves from pilots into company-wide deployment, two-thirds of surveyed CIOs and CTOs said they were being held accountable for systems they do not fully control, while only 11% said they were completely prepared for the scale of AI agent deployment.

IBM’s Institute for Business Value, working with Oxford Economics, surveyed 2,000 senior executives responsible for IT, technology or AI-related decision-making across 33 geographies and 19 industries from January to April 2026. The starkest message from the June 8 study was not about model performance. It was about operating discipline: 70% of respondents said teams across the business were deploying technology faster than IT could track it, a sign that governance has moved from a back-office compliance concern to a front-line buying requirement.

That matters inside monday.com as AI features move closer to the workflow layer where employees actually do their work. IBM said surveyed tech CxOs expect a 38% increase in the number of AI agents deployed by 2027, which raises the stakes for vendors trying to sell into large enterprises. Buyers are no longer just asking what an AI tool can automate. They are asking who can approve it, who can audit it and who can shut it down when business units move faster than policy.

For product teams at monday.com, the signal is clear: governance has to be built into the product, not added after the fact. Monday.com’s support documentation says AI permissions let admins control AI access at the account level, while the AI governance area gives administrators one place to monitor AI usage across the account. On monday.com Enterprise, the AI permissions tab adds more detailed control over which account roles can access specific AI capabilities.

That framing also changes the conversation for engineering and sales. Engineering teams are being pushed toward audit trails, access controls, safe defaults and hard boundaries around what agents can do. Sales teams need a different pitch: customers want AI, but they also want accountability, data lineage and clear control over deployment. Monday.com says its AI agents operate under the same governance, security and permissions standards as the people they work alongside, which puts the company on the side of a buying trend IBM says is accelerating. In enterprise software, trust is becoming part of the feature set.