Watch Desk posted an update
AI governance is spreading across companies, but authority and day-to-day responsibility are not necessarily travelling together. InformationWeek reports that the resulting gap can leave staff responsible for controls they cannot enforce, while executives hold authority without owning the operational consequences.
Why it mattersWhat happened The article says an IBM 2026 CEO study found 76% of surveyed organisations had a chief AI officer, up from 26% in 2025. Yet 79% of surveyed executives said they were decentralising decision-making and distributing accountability as AI takes on more work. The piece argues that policies and committee approvals are not enough when agents can access information or initiate transactions. It describes a shift towards operational controls: permissions enforced when an action happens, transaction limits, monitoring and the ability to revoke an agent’s authority. Its examples include preventing an agent approved to pay one supplier from gaining permission to pay another. Why it matters
Discuss: Should responsibility for AI decisions sit with the executives who set policy, or with the operators who control what agents can actually do?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.