Top story
Dataiku's new Global AI Confessions Report (900 CEOs, Harris Poll, Feb-Mar 2026) found that 78% of CEOs say AI could now cost them their job and their company's future. The pressure isn't coming from AI failing outright, it's coming from boards that have stopped accepting "we're investing in AI" as a sufficient answer: 62% of CEOs say their board is actively pressuring them for measurable, AI-driven results. The center of gravity has moved from adoption metrics to personal accountability at the top.
Quick hits
- IBM/Oxford Economics (2,000 CEOs, 33 geographies): 76% now have a Chief AI Officer, up from just 26% in 2025, but accountability for AI decisions is decentralizing as fast as the role is spreading.
- Pearl Meyer (108 board members and executives, published in Fortune): 90% of boards say the C-suite owns AI strategy. Inside the C-suite itself, only 32% agree the C-suite as a group is accountable, the rest split across functional heads, business leaders, and the level below.
- Dataiku: 80% of CEOs globally say their job is at risk if AI fails by the end of 2026, up from 74% just a year ago.
- Across all three studies, the open question in 2026 isn't whether AI works, it's who answers for it when it doesn't.
Insight for practice
Most AI governance conversations start with frameworks and risk committees. The data points to a simpler gap: ownership. Boards assume the C-suite has it; the C-suite is split four ways internally on who "it" even is. Before adding another governance layer, leadership teams need one direct conversation: who, by name and title, answers for an AI failure that reaches the board. If that answer takes more than one sentence, the gap is real, and it will surface in public before it surfaces in a committee.