The term "deepfake" refers to AI-generated media, typically video, images, or audio, that replaces or fabricates a real person's likeness. The underlying technology learns patterns from large datasets of real images and can then produce new content that was never photographed or recorded.
A few years ago, creating convincing deepfakes required technical skill and expensive hardware. That is no longer true. Consumer-grade apps can now produce synthetic images of a real person in minutes, using only a handful of source photos. Many of those apps are free.
The most common misuse targeting teenagers involves non-consensual intimate imagery: fake explicit images generated from real, innocuous photos. The images are fictional, but the person depicted is real, and the harm, social, psychological, and sometimes legal, is real.
Deepfakes affecting teenagers show up in a few distinct patterns.
The most damaging involves fake intimate images of classmates, shared in private group chats or posted to anonymous accounts. Victims typically find out secondhand, from a friend who saw the images circulating. The psychological toll is significant: students report anxiety, social withdrawal, and fear of returning to school.
A second pattern involves fabricated audio or video used for harassment, put-on clips designed to humiliate a peer. A third involves impersonation: AI-generated content made to look as if a student said or did something they did not, then shared as though real.
Research from Thorn, a nonprofit focused on child protection technology, found that roughly 1 in 10 minors report knowing of cases where their peers used AI to generate nude images of other students.3 A follow-up Thorn study in 2025 found that 1 in 17 teens have been targets themselves, and that 1 in 8 personally know someone who has been victimized. Meanwhile, a Center for Democracy and Technology survey found that 40 percent of students were aware of deepfakes depicting individuals in their own schools during the 2023-24 school year.4 This is not a future concern. Now it’s on platforms schools do not monitor and in channels that are invisible to most parents.
For most of the period following the Westfield incident, students and families had little federal recourse. That changed in May 2025, when President Trump signed the TAKE IT DOWN Act into law.
4 The Act makes it a federal crime to knowingly publish non-consensual intimate imagery, including AI-generated deepfakes, and requires online platforms to remove flagged content within 48 hours of a victim's notice. Penalties for content involving minors can reach three years' imprisonment.
This is meaningful progress. But federal law addresses what happens after harm has occurred. It does not prevent a student from creating a deepfake in the first place, and it does not resolve the practical questions schools face when these incidents arrive through their doors. The social damage reaches the building on Monday morning well before any legal process begins.
Schools still occupy difficult ground. The content is frequently created off-campus, distributed through personal devices, and circulates on platforms the school has no access to. What schools can and should do in response remains, in many districts, an open question.
There is no single intervention that resolves this. But there are a few things worth doing now.
Students need to understand what deepfakes are before they encounter them. Most teenagers have not been taught that images circulating in their group chats may be entirely fabricated, or that generating those images is now a federal crime. That education belongs in school, and it needs to be specific: here is what the technology does, here is who gets harmed, here is what the law says.
Schools also need updated response protocols. When a student reports that fake images of them are circulating, most existing harassment procedures do not map cleanly onto the situation. Who handles the incident? What authority does the school actually have? Is law enforcement involved? Having those answers ready in advance matters.
For parents, the most useful step is a direct conversation: not a warning, but a real discussion about what this technology is and what their child might see or encounter. Teenagers who understand deepfakes are better equipped to recognize them, to support a peer who has been victimized, and to think twice before participating in sharing them.
We do not believe this problem is unsolvable. But solving it starts with knowing what it is. The students at Westfield High School did not encounter something exotic or theoretical. They encountered tools their classmates downloaded on a phone. That is the scale of the problem, and it calls for a response that matches it.
