Every CEO in America is telling their workforce to adopt AI or get left behind. A new study from Boston Consulting Group says that advice might be backfiring spectacularly.

BCG surveyed 1,488 full-time US workers and published the results in the Harvard Business Review. The finding: 14% of all workers are experiencing what researchers call "AI brain fry," a form of cognitive exhaustion caused by constantly monitoring, correcting, and managing AI tools. Among marketers, that number jumps to 26%. Among software developers, it is even higher.

The irony is almost too perfect. AI was sold as the thing that would eliminate tedious work. Instead, it created a new kind of tedious work: babysitting the AI. Checking its output for hallucinations. Rewriting its mediocre first drafts. Debugging code that almost works but does not quite. The cognitive load did not disappear. It shifted.

Here is the part that should worry executives: the study found that workers experiencing AI brain fry are not just tired. They are less productive. The tool that was supposed to boost output is actively degrading it for a significant chunk of the workforce. And the workers most affected are the ones using AI the most, which tends to be the people companies are relying on to lead their AI transformation.

The BCG study did find one silver lining: when AI handles truly repetitive grunt work, burnout rates actually declined. The problem is not AI itself. The problem is the gap between what AI promises and what it delivers. When you have to spend 30 minutes fixing what the AI spent 10 seconds generating, the math stops working.

This research lands at a critical moment. Companies are in the middle of the largest enterprise AI rollout in history. The assumption driving trillions in spending is that AI makes every worker more productive. BCG just provided hard evidence that the opposite can be true. The question is whether anyone in a corner office is listening.