You feel busier than you have ever been. More tabs, more tools, more outputs. Vibe coding , prompting, using AI tools. Working fast. But your thinking feels slower. You second-guess decisions you would normally make without a pause. You finish the day shattered with nothing you are proud of to point to.

The founders I work with describe the same fog, the kind that clears only after they step away from the screen.

That fog has a name. AI brain fry , the mental fatigue that comes from using, overseeing, or interacting with AI beyond what your brain can take. A Harvard Business Review article said 14% of the AI users surveyed reported it, and some jobs reported it far more than others.

The jobs where AI brain fry happens most often

The article is based on a study by Boston Consulting Group, who surveyed 1,488 full-time US workers and broke the results down by function. Here is the share in each role who reported AI brain fry, highest to lowest. 1. Marketing, 25.9% 2. Human resources and people operations, 19.3% 3. Operations, 17.9% 4. Engineering and software development, 17.8% 5. Finance and accounting, 16.7% 6. IT, 16.0% 7. Sales and business development, 12.5% 8. Customer service and support, 10.6% 9. Service provider and consultant, 10.3% 10. Product management, 8.6% 11. Management and leadership, 8.6% 12. Legal and compliance, 5.6% Marketing fried more than four times the share legal did. If you’re working in marketing, consider that the AI use in your job might mean you need to take more breaks or rework how you’re using AI.

What AI brain fry feels like

Participants described a buzzing in the head, a fog that made clear thought hard, slower decisions, and headaches that sent some of them away from the keyboard to reset. One senior engineering manager in the study summed it up. "What finally snapped me out of it was realizing I was working harder to manage the tools than to actually solve the problem."

Workers with brain fry reported 33% more decision fatigue and far more errors, including a 39% jump in serious mistakes. Their intent to quit rose too, from 25% among those without brain fry to 34% among those with it. The people most fluent with AI are the ones a company can least afford to lose. But they are the most affected.

Why some jobs fry faster than others

Every output you check, every draft you judge, every agent you supervise adds to a load that builds fast. There's a huge mental overload to overseeing robots. Workers with high AI oversight spent 14% more mental effort and reported 19% more information overload. Marketing, HR, and operations sit at the top because those roles create many AI outputs at once across content, people, and process.

Tool-juggling makes it worse. Productivity increases as you add a second AI tool, increases again at a third, then drops once you pass three running at the same time. The jobs at the bottom, like legal at 5.6%, use AI in narrow, slow, high-stakes bursts. Fewer outputs to check means less fog.

Decide what you supervise and what you don't

Not every AI output deserves your full attention. Before you open the next tool, split the work into two piles. Steps that can run on their own, and outputs where your judgment is irreplaceable. Check the second pile once the process is done, and leave the first alone.

Treat your attention as the finite resource it is. A founder running six tools and double-checking each one is spending the most expensive thinking in the business on the cheapest tasks. Put it where only you can add the value.

Make AI a manager-hours job

Split your day by maker and manager time. Some blocks are for deep, uninterrupted creation (maker). Other blocks are for meetings, admin, and supervision. Spinning up agents, checking outputs, and playing with new tools is manager work. Keep it inside your manager hours and your maker hours stay clear for thinking.

Get intentional about gaps. When AI is running and you have two minutes to spare, resist filling it by opening another tab. The task-switch costs you more in re-entry than the time it saved. Instead, breathe deeply and reset your mind. When the AI next needs your input, you'll feel more calm and clear.

The 70% problem starts in your inputs

A common complaint is that AI gets everything to 70% and leaves you to finish the rest by hand. When the tool stops short of perfection, look at what you gave it. A vague prompt produces a vague draft, and then you spend an hour rescuing work that better instructions would have got right the first time.

Get good at prompting before you blame the tool. Give context up front. Who the output is for, what good looks like, your brand, the format, the constraints. Front-load the thinking and the first version comes back closer to done, which leaves you far less to clean up and far less fog by the end of the day.

Stop AI from frying the brain that runs your business

Your brain is the asset every product, pitch, and decision builds on. AI can stretch it or fry it, and the difference comes down to how you use it. Decide what you supervise, keep AI inside your manager hours, sharpen your prompts, and let it take the repetitive load while you keep the judgment. Use AI as an extension of your superpowers, not something that makes you want to quit.

Get my free AI playbook to get your time back as a founder.