The Global School Deepfake Sexual Abuse Crisis: An Unprepared System Faces an Exploding Threat
It almost always begins with a single ordinary photo pulled from a public social media profile. Across the globe, teenage boys are scraping images of their female classmates from Instagram and Snapchat, then running them through notoriously harmful “nudify” AI tools to generate non-consensual fake nude photos and deepfake videos of these girls. These manipulated explicit materials spread like wildfire across entire student bodies, leaving victims reeling from humiliation, a deep sense of violation, hopelessness, and crippling fear that the fake images will haunt them for the rest of their lives.
The school-based deepfake sexual abuse crisis began gradually a few years back, but has exploded in scale ever since, as the AI tools needed to create this explicit content have become far more accessible to the general public. An analysis of publicly reported incidents conducted by WIRED and Indicator — a publication focused on digital deception and misinformation — finds that deepfake sexual abuse has impacted roughly 90 schools worldwide, and harmed more than 600 students to date.
The analysis confirms that since 2023, school students, overwhelmingly high school boys, across at least 28 countries have been accused of using generative AI to create sexualized deepfakes targeting their peers. Because this explicit content features minors, it is legally classified as child sexual abuse material (CSAM) in most jurisdictions. This review is believed to be the first comprehensive look at real-world cases of AI deepfake abuse occurring in secondary schools across the globe.
Taken as a whole, the analysis lays bare the global reach of harmful AI nudification technology — an unregulated industry that generates millions of dollars in annual revenue for its creators — and reveals that in the vast majority of incidents, schools and law enforcement agencies are largely unprepared to respond to these severe acts of sexual abuse.
Since 2023, nearly 30 publicly reported deepfake sexual abuse cases have been documented across North America. These include one incident with more than 60 alleged victims, another where the victim was incorrectly temporarily expelled from school, and multiple cases where students across several different schools were targeted at the same time. More than 10 cases have been publicly reported in South America, over 20 across Europe, and another 12 combined in Australia and East Asia.
Experts widely agree that the true scale of school-based deepfake sexual abuse is far larger than these documented numbers suggest. A survey conducted by UNICEF, the United Nations children’s agency, estimates that 1.2 million children had non-consensual sexual deepfakes created of them in just the last year. One in five young people in Spain told researchers from Save the Children that fake deepfake nudes had been made of them. Child protection organization Thorn found that one in eight teens personally know someone who has been targeted by this abuse, and a 2024 survey by the Center for Democracy and Technology found that 15 percent of responding students were aware of AI-generated deepfake abuse linked to their own school.
“It would be very hard to find a single school today that has not been touched by this issue,” says Lloyd Richardson, director of technology at the Canadian Centre for Child Protection. “The most critical priority right now is figuring out how to support victims when an incident occurs, because the harm caused by these acts can be life-altering.”
WIRED and Indicator’s analysis focused only on incidents that had been publicly reported with specific verifiable details, such as school locations and confirmed victim counts. Most of the reporting included in the analysis is in English, meaning there is a major gap in data from many non-English speaking countries. The vast majority of incidents are never covered by the press, even when they are discovered, and many are handled privately by school administrations and law enforcement without any public disclosure.
Even with these limitations, clear patterns have emerged across reported cases. In nearly every confirmed incident, teenage boys are the ones accused of creating the fake images or videos. The materials are most often shared via social media apps or group instant messaging chats among classmates, and the harm inflicted on victims is profound. “I’m scared that every time anyone looks at me, all they see is those fake photos,” one victim from Iowa shared earlier this year. The family of another victim described, “She cries nonstop. She won’t eat anymore.”
Victims almost uniformly report avoiding school to avoid coming face-to-face with the people who created the explicit deepfakes of them. “She feels completely hopeless because she knows these images will almost certainly end up on public internet forums and be accessed by pedophiles,” says attorney Shane Vogt, who is representing an anonymous New Jersey teenager alongside three Yale Law School students — Catharine Strong, Tony Sjodin, and Suzanne Castillo — in a lawsuit against a nudification AI service. “She is in constant severe distress knowing these images exist online, and that she will have to monitor the internet for the rest of her life to stop them from spreading further.”
In response to the growing risk, schools in South Korea and Australia have begun allowing students to opt out of having their photos included in school yearbooks, or stopped posting individual student photos on their official social media accounts, over concerns the images will be used to create deepfakes. “Across the world, we’ve seen cases where school photos are pulled from public social media pages, altered with AI, and turned into harmful deepfake content,” one Australian school administration explained. “Going forward, school publications will use side profiles, silhouettes, shots from behind the head, wide distant group shots, creative filters, or approved stock photography instead of clear individual student photos.”
AI-generated sexual deepfakes have existed since roughly the end of 2017, but the rise of increasingly powerful generative AI systems has spawned a shadowy unregulated ecosystem of “nudification” or “undress” tools. Dozens of apps, social media bots, and websites now let anyone create sexualized deepfakes of another person in just a few clicks, with no advanced technical knowledge required.
“What AI has changed is the scale, speed, and accessibility of this harm,” says Siddharth Pillai, cofounder and director of the RATI Foundation, a Mumbai-based organization that works to prevent violence against women and children. “The barrier to creating these deepfakes has dropped dramatically, which means more people, including teenagers, can create extremely convincing fakes with almost no effort. Like many AI-fueled harms, this has led to a flood of abusive content that is impossible to fully police.”
Amanda Goharian, director of research and insights at child safety organization Thorn, says the group’s research shows that teenagers who create abusive deepfakes have a wide range of motivations, from sexual interest and simple curiosity to revenge, or even just doing it to complete a dare from peers. Studies of adult creators of non-consensual sexual deepfakes similarly find a wide range of motivations for creating the content. “The goal isn’t always sexual gratification,” Pillai notes. “More and more often, the main intent is humiliation, degradation, and exerting social control over the victim.”
“This is not just a problem with technology,” says Tanya Horeck, a professor of feminist media studies and researcher focused on gender-based violence at Anglia Ruskin University, who has studied sexualized deepfake abuse in UK schools. “This is rooted in long-standing gender power dynamics that create the conditions for these crimes to happen.”
As the number of deepfake incidents in schools has risen in recent years, the response from schools and law enforcement has been wildly inconsistent. Parents have repeatedly complained that officials fail to take meaningful action after an incident is reported. In one case, a school took three full days to notify police of a deepfake abuse incident; in another, a victim reported that the students responsible faced no immediate consequences at all. Sometimes creators face criminal charges for creating and possessing CSAM, while other times they only face school suspension or lesser criminal charges, if any. In March, two Pennsylvania students pleaded guilty in juvenile court to felony CSAM charges for creating deepfake images and videos of 60 female classmates, and were sentenced to just 60 hours of community service.
In case after case, it is teenage girl victims and their families that have led the push for change, often moving much faster than politicians, who have been slow to update laws to address this new harm. Teenagers have organized walkouts to support victims, protested against perpetrators, helped create public online training resources, and successfully pushed for legislative change — including contributing to the passage of the U.S. Take It Down Act, which requires tech platforms to remove non-consensual intimate images within 48 hours of a request. Separately, the United Kingdom and European Union are moving forward with bans on nudification apps, while Australia’s eSafety regulator has already taken enforcement action against several of these services.
“Too often, when children do speak up about abuse they’ve experienced, the response they get is completely inadequate,” says Afrooz Kaviani Johnson, a child protection specialist at UNICEF. “How adults respond when a child discloses abuse has a huge impact on both their recovery and whether they will feel safe coming forward again if something like this happens in the future.”
“There is so much work needed to get schools up to speed on the current threat landscape, student and victim rights, deterrence, policy updates, and crisis preparedness,” says Evan Harris, a former teacher and founder of Pathos Consulting Group, which runs deepfake response training for schools across the U.S. Harris says preparedness can range from teaching students about the harm and illegality of creating explicit deepfakes to helping school administrators learn basic digital forensics and evidence gathering. “It is essential that we give students the tools, vocabulary, and support they need if they experience deepfake abuse or learn about it happening to someone else,” says Robyn Little, senior director of educational digital strategy at McDonogh School in Maryland, which has partnered with Harris on training programs.
Schools’ deepfake problems extend far beyond students creating abusive sexual content of their peers. On multiple occasions, students have created sexually explicit deepfakes of their teachers, placing them in humiliating scenarios or making it appear they said things they never actually said. One school in Oregon was forced to bring in substitute teachers after all its regular staff called out sick to protest a social media account that shared manipulated images of faculty. Another report documented deepfakes that showed teachers getting on their hands and knees to eat dog food, or holding a gun.
“While non-consensual sexual deepfake abuse of students is the most urgent, serious, and damaging of the deepfake risks schools face,” Harris says, “it is just one of half a dozen or more deepfake harms that are emerging at the same time, and schools are particularly vulnerable to all of them. They are largely unprepared because of limited resources, and because they are already stretched so thin by so many other priorities.”
