MIT Study Says ChatGPT Might Be Making You Dumber

Is outsourcing your brain to ChatGPT quietly frying your cognitive circuits? A new study out of MIT’s Media Lab suggests the answer might be yes, especially if you’re leaning on AI to help with your homework.

Researchers at MIT recruited 54 people between 18 and 39 and split them into three groups: one wrote SAT-style essays with ChatGPT, another used Google Search, and the third had to rely on pure brainpower. As they wrote, an EEG scanned each participant’s brain activity. The result? The ChatGPT group’s brain activity was consistently the lowest across 32 different brain regions. The essays got progressively lazier too, often ending in a simple copy-paste job. Meanwhile, their neural engagement flatlined.

MORE: I Tested DeepSeek, This Is How It Compares To ChatGPT

DISCOVER: ChatGPT’s AI Doll Challenge: The Backlash Is Beginning

Image courtesy of medium.com

The study’s lead researcher Nataliya Kosmyna didn’t hold back: “What really motivated me to put it out now before waiting for a full peer review is that I am afraid in 6-8 months, there will be some policymaker who decides, ‘let’s do GPT kindergarten.’ I think that would be absolutely bad and detrimental,” she says. “Developing brains are at the highest risk.”

To be fair, the study hasn’t been peer reviewed yet, and the sample size is small. But the findings tap into a broader question: What happens when we replace curiosity and problem-solving with a chatbot?

Kosmyna and her team had participants write essays on classic SAT prompts like the ethics of philanthropy. Those in the ChatGPT group churned out near-identical responses, full of bland phrasing and repeated ideas. English teachers reviewing the essays called them “soulless.” Brain scans confirmed what the red pens suspected – low executive function and minimal attention.

MORE: China’s Deepseek Could Be Bigger Than ChatGPT

DISCOVER: Chat GPT vs Influencers

And when the tables turned and the AI-reliant group had to rewrite their essays without help, they floundered. Most couldn’t recall what they’d written. Their brains barely lit up on the EEG. “You basically didn’t integrate any of it into your memory networks,” Kosmyna explains.

By contrast, the no-AI group had the highest neural connectivity in the brain’s alpha, theta and delta waves, all of which are associated with creativity, memory and deeper thinking. They also felt proud of their work. Even the Google group held its own, showing strong brain activity and high satisfaction. It seems there’s still something to be said for the humble search bar.

MORE: How to Create Your Own AI Doll with ChatGPT: Step-by-Step Guide to the Viral Barbie Box Trend

DISCOVER: Why Airplane Mode On Flights Actually Matters – According To A Pilot

Psychiatrist Dr Zishan Khan agrees: “From a psychiatric standpoint, I see that overreliance on these LLMs can have unintended psychological and cognitive consequences,” he says.

Ironically, after the study went live, people fed it into ChatGPT to get a summary. Kosmyna had anticipated this and planted “AI traps” to limit what the bots could extract.

Her next study? Testing brain activity in coders with and without AI. Spoiler: “the results are even worse.”