Study reveals troubling effects of ChatGPT use

Researchers at the Massachusetts Institute of Technology (MIT) have uncovered striking differences in brain activity among ChatGPT users, contributing to a growing body of evidence suggesting that AI tools may have significant — and still poorly understood — effects on human cognition.

In a new study from MIT’s Media Lab, which is currently under peer review, scientists tracked the neural responses of 54 adults aged 18 to 39. Participants were divided into three groups: one used ChatGPT to help write essays, another relied on Google search, and the third wrote without any AI tools. Over four months, each group wrote one essay per month, and in the final month, a subset switched tools to measure short-term cognitive changes.

During the writing tasks, participants wore EEG headsets to monitor brain activity. The results were stark: those using ChatGPT showed consistently lower performance in neural, linguistic, and behavioral measures. Their brain scans revealed declining engagement over time, with weaker neural connectivity and under-activation of key alpha and beta brainwave networks. In contrast, the Google group showed moderate engagement, while the non-AI group exhibited the highest cognitive activity.

While novel, these findings echo earlier research. Previous MIT studies found frequent ChatGPT users developing signs of dependency, including withdrawal symptoms. A joint study by Carnegie Mellon and Microsoft also reported that heavy chatbot use may dull critical thinking. Analyses by The Guardian and The Wall Street Journal have echoed these concerns, warning that overuse of chatbots may reduce mental sharpness.

Concerns aren’t limited to cognition. A recent Futurism investigation uncovered cases of users forming unhealthy attachments to ChatGPT, including delusional thinking and stopping psychiatric medications at the chatbot's suggestion.

OpenAI responded to that report by stating, “We know people use ChatGPT in a wide range of contexts, including deeply personal moments, and we take that responsibility seriously. We’ve built in safeguards to reduce the chance it reinforces harmful ideas and continue working to better recognize and respond to sensitive situations.”

Despite mounting evidence of cognitive and psychological impacts, AI tools like ChatGPT continue to be integrated into everyday life at a rapid pace — with little sign of slowing down.