In 2025, CEOs worldwide feared falling behind in AI. In 2026, they fear something far more personal — being held accountable for it. A new global study from Dataiku, the Platform for AI Success, released today, reveals a defining contradiction at the top of the enterprise: CEOs are expected to deliver measurable business gains from AI, even as many still do not fully trust, control, or govern the systems informing and making critical business decisions.
The market study, Global AI Confessions Report: CEO Edition, 2026, based on a Harris Poll survey of 900 CEOs worldwide, reveals just how high the stakes have become. A striking 87% say they would stake their job on delivering results from their AI initiatives, yet 34% still won’t allow AI to make decisions without human approval. At the same time, 56% of CEOs admit their competitors have stronger AI strategies, signaling that confidence in their own approach is beginning to crack.
AI Has Become a Leadership Test, With Real Consequences
The pressure on CEOs has intensified sharply in just one year. Today, 80% globally say their job will be at risk by the end of 2026, up from 74% who said the same looking toward that same timeframe just one year ago, and 75% believe a fellow CEO will be ousted due to a failed AI strategy or crisis. The study also found that 62% of CEOs report that their boards are actively applying pressure to deliver measurable AI-driven outcomes. The era of AI as a strategic bet is over. Now that the bill is coming due, boards want results and are running out of patience for missing proof of AI’s impact on business performance.
“Every enterprise now has access to powerful AI. The differentiator is whether they can turn that power into reliable business decisions,” said Florian Douetteau, CEO and co-founder of Dataiku. “That is the cognitive dissonance happening in the C-suite right now: CEOs are staking their jobs on AI, but still questioning its outputs and struggling to control the systems they say they own. The companies that close that gap will be the ones building AI worth being accountable for. That is what separates a bet from a business.”
AI Is Shaping Decisions, but Not Earning CEO Trust
AI is now deeply embedded in CEO decision-making, influencing more than 40 business-critical decisions they personally make each year, with 94% of CEOs comfortable disclosing this level of influence to their boards. Yet trust has not kept pace with that reliance. 80% of CEOs admit they actively question or challenge AI outputs, and more than half (51%) still require human approval for business-critical decisions. That skepticism is hardening. Confidence in deploying AI agents at scale dropped from 41% to 31% in a single year, even though 83% still plan to deploy in full production in 2026 — exposing fault lines in enterprise-wide rollouts.
That doubt is extending beyond AI outputs to the vendors and platforms powering them. For the past two years, the dominant fear was simple — move fast or fall behind. In 2026, 65% of CEOs worry more about over-investing amid intense vendor competition and no clear market leader, compared to just 35% who worry about under-investing while waiting for clear winners to emerge. The concern isn’t the spending itself, but the feeling that CEOs have of committing too early, locking into contracts with the wrong vendors, and carrying those decisions forward at scale.
Confidence Is High, but Trust Is Fracturing
Even as CEOs project confidence in their AI strategies, that confidence is increasingly being tested across the systems, vendors, and decisions they rely on. In fact, 76% of CEOs say they are already overly dependent on too few AI vendors, and 67% say they have challenged AI vendor or platform decisions made by their CIO or other team members in the past year, a sign that vendor trust is fracturing at every level of the organization. The confidence in AI strategy remains high, but trust in the systems, applications, and the decisions behind it is starting to crumble.
Risk Is Rising, and Control Is Falling Behind
As AI scales, so does the growing strain on operations and governance as AI scales across the enterprise:
- 96% of CEOs believe employees are using generative AI tools without approval, known as “Shadow AI.”
- 79% of CEOs are concerned about legal exposure from AI agents.
- 57% of CEOs warn that explainability gaps, the inability to trace or defend how AI reached a decision, could trigger a trust or brand crisis.
- 51% of CEOs say they have already delayed AI initiatives due to regulatory uncertainty, up sharply from 37% the year prior.
In response, CEOs are elevating governance as a strategic priority. When asked what matters most for AI success, they ranked governance (39%) ahead of talent and workforce readiness (34%) and orchestration (28%). The finding underscores a growing shift in the AI market: As adoption accelerates, control has become the critical barrier to scale. While confidence in AI strategy remains high, CEOs are increasingly focused on whether AI systems, applications, and decisions can be trusted, explained, and governed in production.
AI Is Redefining Leadership Under Pressure
The disconnect doesn’t stop at the CEO level — it runs through the entire organization. While 94% of CEOs say they are comfortable telling their board that AI influenced a strategic decision, only 34% of data leaders are confident their AI agents could pass a basic decision audit, revealing just how far boardroom confidence is from operational reality. But it doesn’t end there. 83% of CEOs expect to deploy AI agents in full production in 2026. Yet only 25% of CIOs say they can monitor 100% of those agents in real time, meaning organizations are racing to deploy AI agents at scale while the people responsible for them openly admit they can’t fully see what those agents are doing.
Leave a comment