MIT Study Shows ChatGPT Use Cuts Brain Activity by Up to 55%

**TL;DR** Using ChatGPT while writing essays reduced brain activity by up to 55% in a controlled MIT trial, and similar AI reliance has weakened tumor detection skills in doctors.

**Context** Researchers at the MIT Media Lab tested whether outsourcing thinking to AI changes brain function. They built on earlier work showing that search engines can weaken memory, a phenomenon known as the Google effect. The study aimed to see if large language models produce a stronger cognitive offloading effect.

**Key Facts** Fifty‑four students were randomly assigned to three groups: one used ChatGPT, one used Google search with AI summaries off, and one used no technology while writing short essays. Brainwave measurements showed the ChatGPT group’s activity dropped by up to 55% compared to the no‑tech group, indicating less engagement in creativity and information processing areas. After the task, ChatGPT users struggled to recall content from their own essays and reported feeling little ownership of the work. Vivienne Ming warned that treating AI as a default thinking mode harms especially intelligent individuals by eroding independent thought. In a separate longitudinal study, medical professionals who used an AI tool for colon cancer screening over three months performed worse at spotting tumors when the AI was turned off, suggesting skill atrophy when the aid is removed.

**What It Means** The experimental design supports a causal link: using ChatGPT for writing directly reduces neural effort during the task. The doctor study, while observational, shows a similar pattern of degraded performance without AI assistance. Practical takeaways include treating AI as a supplement rather than a replacement for active thinking, setting limits on reliance for routine cognitive work, and periodically practicing tasks without AI to maintain skill. Readers should monitor how frequent AI use affects their own memory and problem‑solving abilities.

Watch for the peer‑reviewed publication of the MIT trial, larger follow‑up studies that track long‑term cognitive changes, and any emerging guidelines on AI use in education and healthcare.