Is AI Dulling Our Mi

The short answer: brain-imaging research suggests that when people accept AI's good answers without effort, they use their own knowledge and deep thinking less. The evidence is early, though, and the way we use AI seems to matter more than whether we use it.

What the brain studies show. The best-known study is from the MIT Media Lab. Researchers used EEG head caps to record the brain activity of 54 volunteers writing essays with their own knowledge alone, with a search engine, or with ChatGPT.

The ChatGPT group had the lowest brain connectivity, and the group relying on its own knowledge had the highest. The ChatGPT users also felt the least ownership of their work and struggled to recall what they had written, and the effects lingered when they later wrote without help. The authors call this "cognitive debt".

Note that this study used EEG, not a brain scanner. Its authors name functional MRI as the next step. An fMRI preprint on children, published in November 2025, found lower engagement of cognitive-control and attention networks while using ChatGPT.

Beyond the scanner. Other research points the same way:

A study of 666 people found heavier AI use linked to weaker critical thinking, mediated by cognitive offloading, and the effect was strongest in younger users.

Randomized trials with 1,222 participants found AI assistance also reduced persistence on later tasks.

The International AI Safety Report 2026 cites a study in which clinicians' ability to detect tumours without AI had dropped by 6% three months after AI support was introduced.

Productive or dulled? Both can be true. One researcher quoted in TIME notes that when people disengage their own thinking, they may re-engage it elsewhere, for example by spending the saved time on more meaningful tasks.

The risk lies in what happens next. One expert describes an "AI-trust spiral", where more use brings more trust, more offloading and less critical scrutiny.

Design matters. In a high-school trial, classes using an AI that answered freely scored 17% lower on the exam, while classes using an AI that gave only hints showed no measurable difference.

A compilation of the studies concludes that passive use (receiving answers) suppresses brain regions involved in effortful thinking, while active use (directing the tool, being challenged) can maintain or raise engagement.

Caution. The evidence is not settled. Imaging studies so far are small, often preprints, and include one contradictory result. The MIT authors themselves reject the word "dumber". The International AI Safety Report calls research on AI, offloading and critical thinking "nascent".

What to do.
Think first, then consult AI.
Ask it to challenge your reasoning, not to replace it.

Verify its claims before you rely on them.

Keep practising the skills you want to keep, such as writing, calculation and recall.

AI is a strong tool, but a mind needs regular exercise.

Mustapha Bature Sallama
Medical/ Science Communicator,
Private Investigator, Criminal Investigation and Intelligence Analysis,

International Conflict Management and Peacebuilding. (USIP)

[email protected]
+233555275880

References
New Scientist (via Carnegie Mellon University), "AI may blunt our thinking skills": https://www.cmu.edu/dietrich/sds/images/news/zach_new_scientist.pdf

PAUSE: A Privacy-Preserving Self-Reflection Tool for AI-Associated Cognitive Offloading (arXiv), citing Kosmyna et al. (MIT Media Lab, 2025) and Gerlich (2025): https://arxiv.org/pdf/2609.13155

Projet Celsius, "What is the impact of AI on the brain? What does scientific research say?": https://projetcelsius.com/en/blog/ai-impact-on-the-brain/

International AI Safety Report 2026: https://arxiv.org/pdf/2602.21012

TIME, "Is AI Making Our Brains Weaker?" (19 May 2026): https://time.com/article/2026/05/19/is-ai-making-our-brains-weaker/

The Algorithmic Bridge, "How AI Affects Your Brain According to the Studies: A (Very Big) Compilation": https://www.thealgorithmicbridge.com/p/what-the-studies-say-about-how-ai