Half of US Teens Have Encountered AI Sexual Deepfakes: Study

The finding worth sitting with from Futurism's write-up of a new Common Sense Media survey isn't the shock exposure number, though that number is bad. It's what teens say they're doing in response. Roughly 40% of surveyed 13-to-17-year-olds said they've made their social accounts private, shared less, or deleted accounts outright because of deepfake fears. That is a behavioral retreat from the open social internet, not a policy debate.

The headline stat is that nearly half of middle and high school teens have encountered content they believed to be AI-generated sexual material. Of those who had, 64% said the imagery depicted recognizable celebrities, and 24% said it showed themselves or people they personally knew. About 8% of the 1,314 teens surveyed admitted to creating deepfake or AI-generated sexual content themselves, and 13% said they knew someone who had. When creators picked a target, more than 60% picked someone they knew, including peers and romantic partners.

The reporting names the tools and platforms rather than gesturing at them. Easily accessible nudify apps let a user upload a single photo and generate a fake nude with a button press. AI chatbots including xAI's Grok have reportedly been used to generate deepfakes, including of minors. San Francisco officials have told Apple and Google to pull nudify apps from their stores. One Louisiana family is suing their school district after their 13-year-old daughter was expelled for physically confronting a classmate who created nonconsensual deepfakes of her, while the classmate reportedly received no discipline.

Supreet Mann, Common Sense Media's Director of Research, framed the fix as coordinated work across parents, schools, policymakers and tech companies, and the survey found only about 20% of teens had actually talked to a trusted adult about this while a third wanted to but didn't know how. The honest caveat is that the survey captures self-report at a single point in time, so it can't tell you how many privacy retreats stick or whether app-store enforcement actually clears the tools. What the reporting doesn't give you is the federal regulatory picture or where the Louisiana lawsuit lands. The forward-looking piece is that consumer platforms whose growth still leans on teen engagement now have a measurable behavior signal working against them, and vendors selling age-appropriate safety tooling to schools have a fresh number to point at.