AI is getting smarter. You might not be.

You’ve probably heard the adage, “AI won’t replace your job. Someone who knows how to use AI will.” Maybe that’s true. But there’s a catch: the more you rely on AI to think for you, the less practice your brain gets doing it on its own—and like any muscle, it weakens without use. So the very tool that’s supposed to make you more competitive could also be making you less capable.

A study by Carnegie Mellon University and Microsoft researchers found that more reliance on AI reduces critical thinking . It also reduces self-confidence as dependence on the technology rises.

Researchers at the Massachusetts Institute of Technology found that ChatGPT users showed dramatically lower brain engagement , along with weaker retention. Most couldn’t even remember what they had just worked on—and their cognitive engagement stayed low for months afterward.

Younger workers are most at risk, with those ages 18–29 showcasing the highest levels of AI dependence . This high confidence in AI leads to less analytical thinking effort in the critical first years of these workers’ professional lives.

You might think that the scales are balanced: if our brains atrophy a little in the process, maybe that’s okay because at least our work output is better than before, right? Wrong. Organization Science found that while submissions for academic journals have increased since ChatGPT’s launch, writing quality has declined —so the idea that AI uniformly improves writing output simply isn’t true.

All of this research points to one conclusion: Unchecked AI use can accelerate skill decay and slow new skill acquisition – and work output isn’t necessarily better for it. Luckily, there are ways to prevent AI-induced skill atrophy and have an improved work product, too, without swearing off AI forever.

You’ve Already Noticed It In Yourself

It’s probably already happened.

The reflex of typing into the chatbot before you’ve thought about a problem for even ten seconds. Maybe you thought you’d use AI “just to get started.” Or maybe you just couldn’t remember how you’d approach the task before AI existed.

The output is often good enough to keep you coming back. But it misses context. It lacks nuance. It doesn’t understand the stakes. And it’s addictive. Over time, you shift from an active producer to a passive evaluator.

AI saves effort by reducing repetition. But repetition is how expertise gets built.

In fact, 65% of young adults reported thinking AI promotes instant gratification rather than real understanding. There’s a difference between accessing information and processing it. Even worse: 79% worry that it’s making people mentally lazier . Yet use keeps climbing .

This isn’t a technology problem. It’s a habit problem—one you can fix.

Five Ways to Keep Your Brain in the Game

When it’s time to write, draft your own version before you open the chatbot window. The draft can be ugly, but it has to be yours. Only then can you ask AI to pressure-test your thinking by suggesting edits—not replacing your version.

In your area of expertise, your judgment is still more valuable than AI’s output. Anyone can generate an output with the click of a button now. The scarce skill is knowing what’s actually good.

2. Build “No AI Zones” Into Your Week.

Make a list of everything you do on repeat. Identify which tasks most benefit from AI and which don’t. Protect those that sharpen judgment. If a task builds clarity, communication, prioritization or community, keep it human.

  • Drafting emails. The time saved is minimal, but the engagement lost isn’t. Even when staying off of the chatbots, many email providers have AI-generated auto suggestions. Maybe these make your sentences sound more succinct or swap out words, but it also removes your voice. Experiment with not using it to preserve your unique tone. Your response will be sharper and your relationships stronger for it.
  • Writing presentation scripts. These are human moments. The goal is connection—not perfect polish. Write your own words based on your own thoughts—and mean them. Tell your own stories.
  • Setting your weekly priorities. It’s tempting to ask AI what to focus on. But deciding what matters—and what doesn’t—is a core leadership skill. Outsource it and you’ll lose your grasp on your own work.
  • Giving feedback to your team. Feedback is a relationship, not a task. The moment you hand it off to a chatbot, you’ve lost the thread of what you actually think about your people’s work and growth.

Treat this time as strength training for your brain. The more you do it, the easier it gets.

3. Interrogate Every Output Out Loud.

Don’t trust AI blindly. After every AI response, ask yourself a series of questions:

  • What did it miss?
  • What would I have done differently?
  • What do I really think about the output?
  • What needs to be addressed first?

That friction between your thoughts and the AI output is the whole point. It prevents passive acceptance. Make a note of your answers and integrate them into your work—filling the gaps AI left open.

4. Teach What You’re Tempted to Outsource.

The fastest way to retain a human skill is to explain it to someone else. If there’s something you’d like to save time on, mentor a junior colleague through it rather than handing it to a machine.

This has a double benefit: It keeps the work in human hands while passing along knowledge to the next generation.

5. Build Human Connections AI Can’t Replicate.

Schedule networking conversations that are purely human, keep the chatbots out of initial strategy meetings and show up to networking events and conferences in person rather than taking the virtual option.

These efforts will demand the skills AI erodes: reading a room, holding nuance, bouncing back from an awkward thinking on your feet and building team trust through presence.

AI fluency is now the baseline . But cognitive sovereignty is the new differentiator.

The goal isn’t to use AI less. It’s to use it without surrendering the knowledge, judgment and critical thinking you’ve spent a career building.

Start this week by asking yourself, “What's one decision I can make by thinking it through myself first?”