The Proliferation of Deepfakes Is Profoundly Changing the Way Teenagers Use the Internet

Nearly half of U.S. teens aged 13 to 17 have seen AI-generated sexual content, with 8% creating it themselves, according to a new nationally representative survey from the nonprofit Common Sense Media of 1,314 teenagers. The report found that 64% of teens who saw such content said it depicted recognizable celebrities, while 24% said it showed themselves or someone they knew, and 67% of teens worry about being impacted by deepfakes, driving nearly 40% to make social media private, limit sharing, or delete accounts. “For a good chunk of teens, one, they’re being exposed to this content — but two, they are creating and sharing this content,” said Supreet Mann, Common Sense Media’s director of research and lead author.

AI-generated sexual content is rapidly remaking young people’s lives as they confront sex and dating — and the effects are as sweeping as they are poorly understood. That’s according to a new survey https://www.commonsensemedia.org/research/teens-explicit-content-and-deepfakes from the nonprofit Common Sense Media, in which researchers dug into how youth are interacting with these new technologies. What they found suggests that the proliferation of this kind of content — and the harm it’s already causing — demand urgent attention. The nationally representative survey drew on insights from 1,314 teenagers aged 13 to 17, as well as additional data from interviews and focus groups, and defines AI-generated sexual content as deepfaked nude images or pornographic material of real people, AI images and videos of people depicted in revealing or sexually explicit scenarios, or “any other sexual content made by AI.” It comes as middle and high schools around the country grapple with an explosion of nonconsensual deepfakes https://www.wired.com/story/deepfake-nudify-schools-global-crisis/ of minors that have largely been created and disseminated by their peers. “For a good chunk of teens, one, they’re being exposed to this content — but two, they are creating and sharing this content,” Supreet Mann, Common Sense Media’s director of research and the lead author of the report, told Futurism . “And when they’re doing that, it’s very often of people they know personally or themselves.” The report found that nearly half of high school and middle school-aged teens have seen what they believed to be AI-generated sexual content, with most reporting that they discovered it accidentally, often by way of social media. Of the teens who reported seeing this kind of content, about 64 percent said it depicted recognizable celebrities, while 24 percent shared that the content showed either them or someone they knew in real life. Roughly eight percent of teens, meanwhile, reported creating deepfakes or other AI-generated sexual content themselves, while another 13 percent shared that they knew someone who had generated sexual content using AI. Of that cohort, more than 60 percent noted that the “content depicted themselves or people they know personally, like peers, friends, or current and former romantic partners.” The anxiety around this kind of material has driven some kids to change their behaviors online, the report found. Around 67 percent of surveyed teens shared that they were at least somewhat worried that they could be impacted by deepfakes, and nearly 40 percent have made their social media accounts private, are limiting what they share, or have fully deleted social media. “About two-thirds of teens indicated that they were worried about this, and many of those young people are changing their online behaviors because of that fear… that’s a very interesting tell of how young people are thinking about the content, and they’re thinking about this lack of agency that they have in the space,” said Mann. And while it’s not a bad thing for kids to think critically about their online presence, Mann added, as it stands, the burden is falling heavily on the shoulders of teens themselves to cope with the deepfake influx. “Why are young people being forced to police themselves,” asked the researcher, “when this really shouldn’t be their burden to carry?” It’s easy to see why teens are so frightened. With a photo and a click of a few buttons, easily-accessible “ nudify https://futurism.com/the-byte/instagram-advertises-undress-apps ” apps, which are widely available online and through app stores https://www.wired.com/story/san-francisco-demands-apple-and-google-delete-ai-nudify-apps-from-app-stores/ , have made it trivially easy to artificially undress people without their consent. General-use chatbots — most notably xAI’s Grok https://futurism.com/artificial-intelligence/elon-musk-xai-sues-grok-user-deepfakes — have also been used to create deepfakes, including of minors. Deepfake scandals have rocked towns https://www.404media.co/radnor-high-school-pennsylvania-ai-deepfakes-child-sexual-abuse-material/ across the US, primarily impacting girls https://www.nytimes.com/2024/04/08/technology/deepfake-ai-nudes-westfield-high-school.html , who have described experiencing crushing anxiety and panic attacks, suicidal thoughts, and a deep loss of trust in their peers and social world as a result of the disturbing content. Some of the teenage creators of deepfake porn, some of whom have confessed to generating hundreds https://6abc.com/post/lancaster-pennsylvania-teens-get-probation-using-ai-create-fake-nudes-classmates/18778852/ of the images, have been arrested and criminally charged. AI-generated sexual imagery depicting minors is considered to be illegal child sexual abuse material under federal law. One family in Louisiana sued a school district https://apnews.com/article/school-deepfake-nude-ai-cyberbullying-0ead324241cf390e1a7f3378853f23cb after their 13-year-old daughter — after first attempting to alert school officials — was expelled for physically confronting a classmate who had created and shared nonconsensual deepfakes of her and her friends. The boy who generated the images received no discipline, the lawsuit alleged. One bright spot from the survey, Mann noted, was that many kids shared that they want to speak with trusted adults in their lives about this strange new world, but aren’t quite sure how to. “Only one in five teens said that they had ever talked to a trusted adult about this. But a third of teens said that they want to talk to someone, they just don’t know how to have the conversation,” said Mann. Younger teens, she added, are more likely to want to talk with trusted adults about this kind of material than older kids are — while one in ten teens shared that they would be most comfortable talking to a chatbot about what they were seeing. To Mann, those responses signal that “kids are open to having these conversations,” and that “getting in there early and creating that space for conversation is really key.” There’s a “unique window of opportunity for parents to start having these conversations with their kiddos, and to just not wait on the conversation,” Mann added. “We we don’t want our kids to be turning to AI chatbots to have conversations, or to not have these conversations at all.” In addition to parents, schools, policymakers, and tech companies — the latter of which are building the tools being used to generate this material, and run platforms where young people are encountering AI-generated sexual content — need to be proactive about understanding this kind of material and mitigating its harms. One challenge is to limit the creation of deepfakes to begin with, an effort that has some local officials looking at app stores as a possible chokepoint in the deepfake supply chain: last week, the city of San Francisco commanded https://www.wired.com/story/san-francisco-demands-apple-and-google-delete-ai-nudify-apps-from-app-stores/ Apple and Google to remove “nudify” apps from their respective app stores. “This is really a this is something that needs to be tackled from all angles,” said Mann. In the meantime, it’s kids who are confronting the lived impact of artificial porn and deepfakes in their personal and social lives. “We grew as AI grew. We grew up together. Adults don’t have the same life experience that we do,” one 17-year-old girl told researchers. “They don’t understand how advanced and how intense AI is.” More on deepfakes: Elon Musk’s xAI, Which is Being Sued Over AI-Generated Sexual Deepfakes, Sues Grok User Over AI-Generated Sexual Deepfakes https://futurism.com/artificial-intelligence/elon-musk-xai-sues-grok-user-deepfakes