AI is supposed to make work easier. But new research suggests it might be doing the opposite.
The idea behind AI at work is simple. Automate tasks, reduce effort, and free up mental space. But in reality, many workers are experiencing what researchers now call “brain drain,” where using AI actually increases cognitive load instead of reducing it.
The reason comes down to how the brain works.
Humans are not built to juggle too many streams of information at once. Even what we call multitasking is limited. As research shows, “our brains are poor jugglers of information,” especially when constantly switching between tasks and inputs.
And AI adds more layers to manage, not fewer.
Instead of doing one task, workers now have to prompt AI, review its output, correct mistakes, and decide what to trust. That oversight creates a new kind of workload, one that is mentally heavier even if it looks more efficient on paper.
In some cases, it is even measurable.
Studies show that high levels of AI oversight can increase mental effort, fatigue, and information overload, leading to what some describe as “brain fry,” a mix of mental fog, slower thinking, and more errors.
There is also a deeper trade-off.
AI is removing repetitive tasks, but those same tasks used to give the brain time to recover. Without them, people stay in constant high-focus mode, which drains mental energy faster than expected.
That is where the concept of cognitive load becomes important.
It is not just about how much work you do, but how mentally demanding that work is. And AI, when poorly used, can increase that demand instead of reducing it.
The response is not to abandon AI, but to rethink how it is used.
Experts suggest creating quiet time without constant prompts or meetings, because insight and creativity often come when the brain is not overloaded. As one key idea puts it, “Eureka moments don’t happen in noisy brains.”
There is also a skill shift happening.
The people who benefit most from AI are not the ones who fully depend on it, but those who guide it. This requires metacognition, the ability to think about your own thinking and use AI as a tool to improve it, not replace it.
What this points to is a subtle shift in how work is evolving.
AI is not just changing what we do. It is changing how much mental energy each task requires.
And that raises a bigger question.
If AI keeps increasing the speed of work but also the mental load behind it, are we becoming more productive, or just more exhausted?
