Microsoft CEO Satya Nadella calls for employee‑style governance of AI agents

Microsoft chief executive Satya Nadella said that as AI agents become capable of handling multi‑step workplace tasks, companies should manage them like employees, assigning each a distinct identity, specific permissions and clear oversight. He argued that such governance is needed to avoid errors such as misinterpreted instructions or accidental data loss.

Nadella highlighted the need for “security, containment, manageability and observability” when deploying autonomous assistants. Microsoft already uses internal tools such as the Agent 365 suite, which combines Entra for digital identity and network access with Purview for data labeling and tracking, to monitor the hundreds of AI coding agents it runs. The comments came on the "Possible Podcast" and reflect a broader industry shift, with firms like Morgan Stanley beginning to grant AI agents access to proprietary data.

The executive’s remarks point to a growing focus on formalizing AI‑agent oversight to mitigate operational risks as these agents move beyond simple chat functions toward more autonomous roles.