Fear of missing the AI wave is pushing corporate boards into hasty decisions, and chief executives are increasingly saying so. A new Boston Consulting Group survey finds that 61% of CEOs think their boards are moving too fast on AI adoption, exposing a widening gap between strategic caution and boardroom urgency.
In BCG’s ”Split Decisions: The BCG CEOs and Boards Survey,” 625 senior leaders were surveyed worldwide, including 351 CEOs and 274 board members at companies with annual revenue of at least $100 million. The message is blunt: AI enthusiasm is colliding with governance, risk management, and the old boardroom habit of mistaking motion for progress.
There is also a familiar corporate pattern here. New technology cycles often create a confidence gap: leaders closest to the day-to-day mechanics of a business tend to see the limits first, while boards worry more about being late than being wrong. AI is amplifying that tension because the vendors are loud, the promises are huge, and the downside of a bad implementation can be expensive without looking dramatic until later.
Boards say they understand AI, CEOs are unconvinced
BCG says 75% of board members believe they understand the substance and value of AI implementation. CEOs are much less impressed. More than a third, 35%, think boards overestimate how easily AI can replace people in everyday workflows, and nearly 40% say boards do not have a realistic picture of how AI changes growth strategy.
That mismatch matters because it affects what gets funded, what gets rushed, and what gets dressed up as transformation. If a board thinks AI is mainly a staffing shortcut, it may push for faster deployment than the business can absorb. If the CEO sees AI as a longer operational redesign, the same plan can look reckless rather than bold.
The strange confidence gap inside the boardroom
The survey also points to a more uncomfortable pattern: the less confident board members are in their own AI knowledge, the more likely they are to believe the company is moving too slowly. That is a neat recipe for bad timing. It encourages urgency without necessarily improving judgment.
BCG partner Julie Bedard argues that the fix starts with direct CEO-led education, not abstract briefings. In practice, that means showing boards what AI tools can really do, where they fail, and which tasks they can augment rather than replace. It is a more useful exercise than letting everyone nod through another slide deck about ”transformation.”
CEO performance is now tied to AI results
The pressure is not coming only from inside the boardroom. CEOs say roughly 35% of their own performance assessment is now linked to AI deployment and returns from AI projects. That changes incentives fast. Once executive scorecards are tied to AI spending, caution can start looking like weakness.
Still, there is broad agreement on one point: future board members will need to understand how AI reshapes their industries. About 80% of respondents said as much. The problem is that agreement on the goal does not mean agreement on the pace, and pace is where expensive mistakes usually hide.
BCG senior partner Judith Wallenstein says companies need to involve boards in AI training more systematically, but also more quickly and with less hand-waving. That is probably right. The next phase of corporate AI adoption will not be decided by who sounds the most excited at a board meeting, but by who can separate useful automation from expensive theater.