MIT Study Finds AI Can Plant False Memories People Truly Believe

An experiment by MIT Media Lab professor Pattie Maes showed that among 200 people who viewed AI-edited photos, the chance of forming false memories rose sharply, and they felt even more confident about them. Other studies suggest AI images and dialogue can plant events that never happened into people's minds. In another test, subjects who wrote essays with ChatGPT showed the weakest brain connectivity and struggled to recall their own words. A 2025 Lancet study found doctors who routinely used AI saw their ability to independently detect precancerous lesions drop by about 20%. Researchers worry people will gradually stop practicing skills machines can handle, a trend called "skill decay." Georgetown research further found that misleading AI summaries can distort the memory people rely on to catch AI errors, so human oversight may fail as a safeguard. This research is still at an early stage.