You bought the AI tools to free up time. The result? Your most engaged employees are working more, making more mistakes, and thinking about quitting.
That’s not an irony. It’s a pattern now documented in research.
”You don’t work less - you just work more”
UC Berkeley spent eight months inside a 200-person tech company as employees genuinely embraced AI. Nobody was pressured. No new targets were set. People used the tools because the tools worked.[1]
The result wasn’t that people freed up time. It was that their to-do lists expanded to fill every hour AI freed up - then kept going.
One engineer put it plainly: “You had thought that maybe, because you could be more productive with AI, you save some time, you can work less. But then really, you don’t work less. You just work the same amount or even more.”[1]
A comment on Hacker News captured the same experience: “Since my team has jumped into an AI everything working style, expectations have tripled, stress has tripled and actual productivity has only gone up by maybe 10%. It feels like leadership is putting immense pressure on everyone to prove their investment in AI is worth it.”[1]
14% experience what researchers now call “AI brain fry”
Harvard Business Review published a study in March 2026 of 1,488 full-time US workers.[2] Researchers asked whether employees had experienced mental fatigue from intensive AI use: difficulty focusing, slower decision-making, headaches, a sense of mental fog.
14% said yes. They call it “AI brain fry.”
A senior engineering manager described it this way: “I had one tool helping me weigh technical decisions, another spitting out drafts and summaries, and I kept bouncing between them, double-checking every little thing. But instead of moving faster, my brain just started to feel cluttered. Not physically tired, just… crowded. It was like I had a dozen browser tabs open in my head, all fighting for attention. What finally snapped me out of it was realizing I was working harder to manage the tools than to actually solve the problem.”[2]
The cost isn’t abstract:
- Those reporting AI brain fry make 39% more serious errors than those who don’t[2]
- They make 11% more minor errors that can be caught and corrected[2]
- They show 33% more decision fatigue[2]
- Among those without AI brain fry, 25% actively intend to quit. Among those with it: 34%[2]
That’s a 39% increase in intent to leave among your heaviest AI users. The ones who should be your most valuable.
Why is this happening?
The research points to oversight as the heaviest load. When an employee no longer just does the work but actively monitors AI agents doing the work, mental effort increases by 14%, mental fatigue by 12%, and sense of information overload by 19%.[2]
Using three or more AI tools simultaneously reduces productivity compared to one or two.[2] Expecting people to accomplish more because AI makes them faster increases mental fatigue by 12%.[2]
The signals matter. “We’re 30% more productive with AI” gets read by employees as: 30% more is now expected of me. In the study, employees at organizations perceived to value work-life balance had 28% lower mental fatigue scores.[2] The messages you send count.
Jason Lemkin, who runs SaaStr and manages 20+ AI agents in practice, describes the cost concretely: “Every minute you spend with them requires active thinking. You’re analyzing output patterns, refining prompts, evaluating quality, making decisions about training data. It’s so many brain cells. It’s exhausting in a completely different way.”[3]
What actually seems to work
There’s an important distinction in the same research. AI use that replaces repetitive tasks - things people didn’t want to do anyway - reduces burnout by 15%.[2] Employees who use AI to get rid of tedious work and then apply the freed time to things they find meaningful: they do better, are more engaged, stay longer.
The problem is most organizations don’t steer toward that. They buy AI tools, measure token usage or lines of generated code, and let pressure to show ROI push people to take on more rather than work differently.
The researchers’ recommendation: 70% of AI transformation should be about people and processes. Not about which tools you buy.[2]
What this means for leaders
Three things you can do now:
Be explicit about what AI is not supposed to mean. If your employees believe AI productivity = more work in the same time, you’re creating exactly the environment that leads to brain fry. Say out loud what you intended.
Look at how people actually use the tools. Are they using AI to replace dull tasks and free up time for more important things? Or have they acquired more AI agents to supervise on top of everything they already had?
Measure outcomes, not activity. Number of AI prompts, token consumption, generated lines of code - these are activity metrics. They say nothing about decision quality, reduced error rates, or whether people actually want to keep working for you.
What makes this hard is that the problem shows up late. Cognitive fatigue builds gradually. The errors start appearing weeks after workload increased. The employee thinking about leaving does it quietly for a while before they tell you.
You can keep pushing and hope it works out. Or you can decide now what kind of organization you want to build with AI.
The organizations that handle it best, according to the research, share one thing: they treat AI as a collective capability with clear norms, not as individual performance to measure and reward.[2]
That’s not a technology question. It’s a management question.
Sources:
- Ranganathan, A. & Ye, X.M. (2026). AI Doesn’t Reduce Work - It Intensifies It. Harvard Business Review, February 9, 2026. Reported by Loizos, C. in TechCrunch: https://techcrunch.com/2026/02/09/the-first-signs-of-burnout-are-coming-from-the-people-who-embrace-ai-the-most/
- Harvard Business Review Research Team. (2026). When Using AI Leads to “Brain Fry”. Harvard Business Review, March 2026. https://hbr.org/2026/03/when-using-ai-leads-to-brain-fry
- Lemkin, J. (2026). Right Now, Managing AI Agents is About as Much Work as Managing Humans. Just Different Work. SaaStr. https://www.saastr.com/right-now-managing-ai-agents-is-about-as-much-work-as-managing-humans/
