Artificial intelligence is widely seen as the engine of a new productivity era. Yet a recent analysis discussed in the Harvard Business Review suggests a more complex picture. The research identifies a growing phenomenon among knowledge workers that the authors describe as “AI Brain Fry” — a form of mental fatigue caused by intensive interaction with, and supervision of, AI tools.
The study draws on a survey of 1,488 full-time employees in the United States who regularly use AI at work. These findings were complemented by qualitative accounts from organisations where multiple AI systems are used simultaneously. The result: roughly 14 per cent of respondents reported experiencing clear symptoms of cognitive overload directly linked to their use of AI.
AI Brain Fry differs from traditional burnout. Burnout is typically associated with prolonged overwork, emotional exhaustion and declining engagement. AI Brain Fry, by contrast, appears as a more immediate form of mental overload. Workers describe sensations of “mental fog”, difficulty concentrating and a persistent buzzing in the head. Decision-making slows, thoughts feel less structured, and many report needing to step away from screens simply to regain clarity.
The pattern is most pronounced in roles that rely heavily on generative AI. Marketing, software engineering, HR, operations and IT show particularly high rates of reported fatigue. In some marketing teams, more than a quarter of respondents indicated experiencing such symptoms. Legal roles, by contrast, appear less affected on average.
The cause lies less in AI itself than in the way it is used. Many knowledge workers now operate several AI systems at once: one for research, another for drafting text, another for coding or data analysis. Increasingly, the nature of work shifts into a supervisory layer. Rather than solving problems directly, employees must evaluate outputs, compare alternatives, correct inaccuracies and determine which suggestions are trustworthy.
This continuous oversight creates a new form of cognitive load. The study finds a clear link between heavy AI supervision and mental fatigue. Roles with the highest oversight demands reported about twelve per cent greater mental exhaustion and significantly higher levels of decision fatigue.
At the same time, a paradox emerges. AI often increases productivity, but it can also expand the overall scope of work. Many employees take on additional responsibilities because tasks can be completed faster with AI assistance. Over time, this leads to work intensification, with some describing the feeling of effectively performing the workload of two or three people.
Another contributing factor is the erosion of boundaries between work and rest. AI-driven tasks can easily be triggered “just for a moment”, while reports and analyses continue to run in the background. As a result, breaks lose part of their restorative function.
The organisational consequences are measurable. According to the study, workers experiencing AI Brain Fry commit around eleven per cent more minor mistakes and nearly forty per cent more major ones. At the same time, their likelihood of considering leaving their job rises noticeably.
Importantly, the research does not suggest that AI inevitably leads to overload. The determining factor is how the technology is implemented. When AI primarily removes repetitive or administrative tasks, burnout levels fall and employees report higher engagement. The problems arise when AI creates additional streams of work or turns employees into constant overseers of automated systems.
For organisations, the implication is clear. AI can increase productivity — but only if its use is designed carefully. Too many tools, unclear responsibilities or incentive systems that reward constant AI use may increase cognitive strain rather than reduce it.
In this sense, AI Brain Fry is less a technological issue than an organisational one. The real challenge is not adopting AI, but integrating it in ways that genuinely lighten the cognitive load of work.
