AI Tools Are Making Your Brain Lazy Without You Noticing

The Shortcut That Rewires You

Here is a quiet confession from the year 2025: we are outsourcing our thinking to machines, and the machines are winning.

Michael Gerlich, a researcher at the University of Applied Sciences in Switzerland, recently published a study that should make anyone who uses ChatGPT, Grammarly, or even Google Maps sit up straight. The paper, titled "AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking," surveyed 666 people across age groups and educational backgrounds. The finding was stark: the more you lean on AI tools, the worse your critical thinking gets (Gerlich, 2025).

Not because the tools are bad. Because they are too good.

The study zeroed in on a concept called cognitive offloading. This is the brain's natural tendency to dump mental work onto external tools. A calculator offloads arithmetic. A GPS offloads navigation. A notes app offloads memory. These are ancient strategies. What changed is the scale. AI tools now offload reasoning itself. They generate arguments, summarize texts, write emails, even simulate empathy. And the more we let them, the less we practice the muscle of thinking through something ourselves.

Gerlich's data showed a significant negative correlation between frequent AI tool usage and critical thinking scores, mediated by increased cognitive offloading (Gerlich, 2025). In plain language: people who used AI tools the most also scored lowest on tests of critical thinking. And the mechanism was clear. They had handed over the work of thinking to the machine. The machine did it. Their brains did not.

This is not a technology story. It is a biology story.

The Young Are Losing the Most

One of the most unsettling findings in Gerlich's study involved age. Younger participants, those under 30, showed higher dependence on AI tools and lower critical thinking scores compared to older participants (Gerlich, 2025). This was not a small difference. It was statistically significant across multiple measures.

Think about what that means. The generation that grew up with smartphones, social media, and now generative AI is the generation most likely to hand over cognitive work to machines. They are not lazy. They are adapted. They have been trained by their environment to expect instant answers. And AI delivers. But the cost is invisible. You do not feel your critical thinking eroding. You only notice that you used to enjoy puzzling through a problem, and now you just ask the chatbot.

The older participants in the study, those over 50, used AI tools less frequently and scored higher on critical thinking tests. This is not because older brains are better. It is because older brains have had decades of practice doing the thinking themselves. They built the muscle before the shortcut existed.

Gerlich's interviews revealed something else. Many younger participants expressed confidence that AI tools were making them smarter. They felt more productive, more informed. But the test scores told a different story. The feeling of fluency, the ease of getting an answer, was being mistaken for understanding (Gerlich, 2025).

This is the core paradox of AI tools. They make you feel brilliant while you are becoming less so.

How the Study Was Done

The methodology matters here because the claim is large. Gerlich used a mixed method approach. First, a survey of 666 participants, measuring AI tool usage frequency, cognitive offloading tendencies, and critical thinking ability. The critical thinking assessment was not a multiple choice quiz. It was a validated test that measures how well people evaluate arguments, recognize assumptions, and draw logical conclusions.

Then came the interviews. Gerlich conducted in depth interviews with a subset of participants to understand the qualitative experience. What did people feel when they used AI tools? Did they trust the outputs? Did they verify them? Did they feel their own thinking had changed?

The quantitative analysis used ANOVA and correlation analysis, standard tools for this kind of work. The qualitative data went through thematic analysis, where patterns in the interviews were identified and coded.

The result was a clear picture. The more people offloaded cognitive work to AI, the lower their critical thinking scores. And this held true even when controlling for education level, age, and prior familiarity with technology.

Higher educational attainment was associated with better critical thinking skills, regardless of AI usage (Gerlich, 2025). This is an important nuance. Education protects you, but only up to a point. If you have a strong foundation in critical thinking, you can use AI tools without losing your edge. The problem is for people who never built that foundation, or who are building it now while simultaneously outsourcing the work.

The Cognitive Offloading Trap

Cognitive offloading is not inherently bad. Your brain has limited working memory. Offloading routine tasks frees up mental resources for harder problems. This is why you write down a phone number instead of trying to hold it in your head. It is why you use a calculator for arithmetic.

But there is a threshold. When you offload reasoning itself, you stop practicing reasoning. And reasoning is a skill. It requires repetition. It requires the uncomfortable feeling of not knowing the answer and having to figure it out.

AI tools remove that discomfort. They make the hard parts of thinking invisible. You never wrestle with a confusing paragraph because the AI summarizes it. You never struggle to articulate an idea because the AI writes the sentence for you. You never have to hold two contradictory ideas in your head and decide which one is true, because the AI gives you a single plausible answer.

This is the trap. The AI is not wrong. It is often right. But rightness is not the same as understanding. And understanding is what critical thinking is built on.

Gerlich's study found that cognitive offloading was the mediating factor. It was not just that AI users were lazy. It was that they had developed a habit of handing over the cognitive work. The more they did it, the less they practiced the skill. And the less they practiced, the worse they got (Gerlich, 2025).

This is a feedback loop. And it is silent. You do not feel yourself getting dumber. You feel yourself getting faster. But speed is not depth.

What the Study Does Not Prove

It is important to be precise about what this research does and does not show.

The study is correlational. It does not prove that AI tools cause a decline in critical thinking. It is possible that people with weaker critical thinking skills are more drawn to AI tools in the first place. The causation could run the other way, or it could be bidirectional. People who struggle with thinking might seek out tools that think for them, and those tools then further weaken their skills.

Gerlich acknowledges this limitation. The study design does not allow for a clean causal claim. What it does is establish a strong association, and it provides a plausible mechanism: cognitive offloading. The interviews add depth. People described feeling less inclined to think through problems on their own. They described trusting AI outputs without verification. They described a shift in their own habits.

But the study also raises an interesting open question. Could AI tools be designed to strengthen critical thinking instead of weakening it? Imagine a tool that does not give you the answer but asks you questions. Imagine a tool that forces you to articulate your reasoning before it responds. That is not the current design of most AI tools. But it could be.

The study does not prove that all AI use is bad. It proves that the current pattern of use, where the AI does the thinking and the human accepts the output, is associated with lower critical thinking. That is a design problem, not a technology problem.

The Education Gap

One of the more hopeful findings in Gerlich's study is that education matters. Participants with higher educational attainment scored better on critical thinking tests, regardless of how much they used AI tools (Gerlich, 2025).

This suggests that a strong foundation in critical thinking is protective. If you have learned how to evaluate arguments, question assumptions, and reason through problems, you can use AI tools without losing those skills. The danger is greatest for people who are still building those skills, or who never built them at all.

This has direct implications for education. Schools cannot ignore AI. Students are using it. The question is whether they use it as a crutch or as a tool. Gerlich's research suggests that the default behavior is to use it as a crutch. The cognitive offloading is automatic. Students ask the AI for an answer, get it, and move on. They do not stop to evaluate. They do not check the reasoning. They do not ask whether the answer is actually correct.

The solution is not to ban AI. The solution is to teach students how to use it without offloading their thinking. This means assignments that require verification. It means asking students to explain why an AI's answer is right or wrong. It means building the habit of metacognition, thinking about your own thinking, even when a machine is doing some of the work for you.

The Productivity Paradox

There is a seductive argument that AI tools make us more productive. And they do. You can write an email in seconds. You can summarize a long document in a paragraph. You can generate a draft of a report without staring at a blank page.

But productivity is not the same as effectiveness. And effectiveness is not the same as understanding.

The interviews in Gerlich's study captured this tension. Participants described feeling more productive, but also feeling less engaged with their own work. They felt like they were producing more, but thinking less. They felt like they were moving faster, but learning less.

This is the productivity paradox of AI. You get more done, but you understand less of what you have done. You produce more output, but you have less insight into how that output was generated. You feel efficient, but you are not building the skills that make you effective over the long term.

The research suggests that this paradox is real. The quantitative data showed that higher AI usage was associated with lower critical thinking scores. The qualitative data showed that people felt the tradeoff, even if they did not name it. They knew they were offloading. They just did not see the cost.

What This Actually Means

The study from Gerlich is not a warning to stop using AI tools. It is a warning to use them differently. Here is what the research actually means for how you should think about your own brain.

• ▸Treat AI outputs as a starting point, not an ending point. Every time you accept an AI answer without evaluating it, you are practicing acceptance, not thinking. The habit of verification is the habit of critical thinking. Do not skip it.
• ▸Use AI to generate options, not conclusions. The best way to use a language model is to ask it for multiple perspectives, then decide which one is strongest. That forces you to evaluate. That is thinking.
• ▸Watch for the feeling of fluency. If an AI answer feels perfectly clear and easy, that is a red flag. Real understanding often feels uncomfortable. You should be able to explain why the answer is correct, not just recognize that it sounds right.
• ▸Build your foundation before you offload. If you are still learning a domain, do not use AI to skip the hard parts. The hard parts are where the learning happens. Use AI only after you have struggled through the problem yourself.
• ▸Check your own confidence. The study found that frequent AI users felt smarter even as their test scores declined. Do not trust your feeling of competence. Trust your ability to produce an answer from scratch, without help. That is the real measure.

The machines are not making us stupid. They are making us comfortable. And comfort is the enemy of growth.