AI is rotting our brains—and the cognitive cost could reach the exam room
Industry Buzz
“AI offers the ability to learn in a way that is most personalized to a person’s learning style. Pacing in learning is important, especially for learners with ADHD. [But] part of learning involves overcoming certain mental obstacles, and AI eliminates opportunities for critical thinking.” — Ryan Sultan, MD, psychiatrist at Integrative Psych
“The real concern isn't that AI diminishes cognitive capacity, but rather how we're using it—as a replacement rather than an amplifier for human thinking.” — Sabrina H. Habib, PhD, chair of Visual Communications at the University of South Carolina
Tools like ChatGPT and other AI chatbots have become ubiquitous in the past few years, and their use has raised an important question: Is it making us dumber?
Professors say they’ve seen a gradual decline in the quality of students’ written work—yet some professors themselves are even using it, leading students to push back.[][] Other people are using ChatGPT to summarize articles, papers, and research instead of reading.[]
Even doctors are using AI for previously unheard of tasks—12% of your physician-colleagues reported using AI for "assistive diagnosis."[] (Overall, 66% of today's physicians are using AI for work-related tasks.)
A 2025 MIT study focused on whether using large language models—aka, AI chatbots—negatively impacts the way humans learn. Their findings? It sure does.[]
Related: AI-induced psychosis is a growing threat. Here's how to recognize it in patients—and how to helpMore about the study
The study divided 54 participants into three groups: A search engine group, a brain-only group, and an AI group. Each group completed three essay writing sessions.
In a fourth session, 18 of the participants were swapped; the AI group was put into a brain-only group to redo the same task, and the brain-only users were reassigned to an AI group.
The researchers used electroencephalography (EEG) to assess cognitive load during the essay writing, and then analyzed the essays using natural language processing (an AI tool—ironically) to judge the writing. Human teachers also helped.
The brain-only participants exhibited the strongest, most distributed cognitive networks. In contrast, the search engine users showed moderate engagement, and AI users displayed the weakest connectivity.
In session 4, AI-to-brain participants showed reduced alpha and beta connectivity (or under-engagement), while brain-to-AI users exhibited higher memory recall and activation of occipitoparietal and prefrontal areas.
Self-reported ownership of essays was the lowest in the AI group and the highest in the brain-only group. AI users couldn’t even quote their own work. Moreover, the AI users consistently underperformed at neural, linguistic, and behavioral levels.
The study’s authors put it simply: “These results raise concerns about the long-term educational implications of [AI] reliance and underscore the need for deeper inquiry into AI's role in learning.”
What do experts think?
The issue of AI isn’t so black and white, it seems. AI might even offer benefits to learning, in the right context, especially if we’re aware of our deficits to begin with. Marlee Strawn, co-founder and head of education at Scholar Education, says she’s seen firsthand in the classroom how learning changes when the new tools enter the scene: “AI isn’t inherently making us less capable thinkers. It’s revealing where we’ve overvalued memorization and undervalued metacognition."
Ryan Sultan, MD, a psychiatrist at Integrative Psych, says AI might be beneficial for certain kinds of learners.
“People learn in different ways as well. One of the biggest barriers to effective learning is learning in a style that’s different from your own. AI offers the ability to learn in a way that is most personalized to a person’s learning style. Pacing in learning is important, especially for learners with ADHD,” Dr. Sultan says.
But there’s a catch: Part of learning involves overcoming certain mental obstacles, and AI eliminates opportunities for critical thinking, Dr. Sultan explains. "Similar to any other muscle in the body, when you do not work or ‘exercise’ the brain, muscle atrophy can occur, and using AI to solve challenging, thought-provoking problems can potentially lead to weakening of the brain’s muscle mass," he says.
Studies have even associated overreliance on AI with worsening cognition, aka "digital dementia." Digital dementia is the "cognitive decline associated with excessive reliance on digital devices [...] potentially exacerbating attention deficit hyperactivity disorder (ADHD) and memory loss," per authors of a 2024 review in Cureus.[]
Sabrina H. Habib, PhD, chair of Visual Communications at the University of South Carolina, says the goal should be to design learning and professional environments where AI enhances—rather than replaces—the cognitive processes needed for deep learning and creative problem-solving.
“This requires intentional integration strategies from educators, professionals, and individuals to preserve human input and agency while using AI's capabilities as a tool rather than a replacement for critical thinking,” Dr. Habib says.
Related: Are your patients diagnosing themselves with AI?