Thirteen percent of K-12 principals reported deepfake bullying incidents in the last two school years — 22% at the high school level. Forty percent of students say they're aware of deepfakes depicting people at their school. The incidents range from fake TikTok accounts with racist slurs to AI-generated explicit videos of teachers made from school website photos.

1. This Is Destroying Teachers (NEA, International Education Union, Victims)

Teachers are developing PTSD, quitting their jobs, and watching fabricated pornography of themselves go viral — made by their own students.

Deepfake victims report anxiety, depression, humiliation, helplessness, and PTSD. The harm doesn't end when the video comes down — victims suffer recurring traumatization as content resurfaces across platforms and follows them into new jobs and new cities. The teaching profession was already in a mental health crisis before AI tools put deepfake creation in the hands of middle schoolers.

The cases are getting worse, not better. In the UK, a teacher resigned after an 8-year-old student used school website photos to create a sexually explicit deepfake of her and two colleagues. The school refused to suspend the child despite the teachers' union advising the act was illegal. She quit a year later, feeling isolated and unsupported. In Pennsylvania, eighth graders created 22 fake TikTok accounts impersonating teachers with racist, homophobic, and sexually explicit content. In Texas, a student was arrested for creating AI deepfake porn of teachers using campus photos.

The International Education Union has formally condemned the trend, providing support to multiple teachers victimized by AI-generated pornography created by their students. The NEA calls deepfakes the "next iteration of bullying."

2. Kids Have Always Done This (Digital Literacy Advocates, Some Parents)

Spitballs became Rickrolls became deepfakes. The impulse is the same — the tools just got scarier.

Students have been pranking and mocking teachers since schools existed. The bathroom-wall caricature of the math teacher, the prank call to the principal's office, the note passed around class. Digital tools added virtual backgrounds and smart board hijacks during the pandemic. The argument: kids testing boundaries with authority figures is a developmental constant. The medium changes; the behavior doesn't.

There's a version of this argument that isn't dismissive. Recognizing the continuity of the impulse matters for designing interventions. If you treat every 13-year-old who makes a dumb video as a criminal, you'll fill juvenile courts without changing the underlying behavior. South Carolina and Florida have built "proportional penalties" into their deepfake laws — factoring in age, intent, and prior history to avoid excessive juvenile charges. The question isn't whether kids should face consequences. It's whether the consequences should be educational or carceral.

3. AI Made This a Completely Different Problem (RAND, Sameer Hinduja, UNESCO)

A drawing on a bathroom wall doesn't get 100,000 likes. A deepfake does.

The "just a new medium" framing collapses under the scale of the difference. Tools like Viggle AI make deepfake creation accessible to children without any technical skill. The outputs spread globally in hours, get hundreds of thousands of views, and provide what looks like visual evidence — making them harder to dismiss than any rumor or caricature. Unlike graffiti, a deepfake can't be painted over. It resurfaces, gets re-shared, follows the victim into new jobs and new cities.

This is a genuine explosion of terrible, fake content. Reports of AI-generated child sexual abuse material to the National Center for Missing & Exploited Children have surged into the hundreds of thousands. That's not a new medium for an old behavior. That's an explosion. UNESCO frames deepfakes as part of a broader "crisis of knowing" — where trust in visual and audio evidence itself is being undermined.

RAND found the problem is already systemic. One in five high school principals has dealt with a deepfake incident. Yet only 19% of students say their school has explained what deepfakes are. Only 13% know how deepfakes affect victims. More than two-thirds of school staff received no or poor-quality training.

4. The Law Is Scrambling (48 States, TAKE IT DOWN Act, First Amendment Scholars)

Forty-eight states have passed deepfake laws. Most of them don't know what to do with a 12-year-old perpetrator.

Publishing sexually explicit images is illegal now. The TAKE IT DOWN Act, signed in May 2025, made it a federal crime to publish sexually explicit images — real or AI-generated — without consent. It covers both minors and adults, meaning teachers are explicitly protected. Platforms must remove nonconsensual intimate imagery on notice. At least 48 states have also enacted their own deepfake legislation.

But the legal framework still has gaps. Deepfakes receive First Amendment protection as free expression, and schools can only restrict off-campus speech if it "substantially impacts" the school environment. Many state laws don't address explicit deepfakes specifically. And most laws weren't written with 12-year-old perpetrators in mind — the penalties are designed for adults.

Some states are building age-aware systems. South Carolina and Florida have tiered penalty structures that factor in the student's age, intent, and whether it's a first offense. Advocates for proportional justice argue the goal should be reeducation over extreme charges. Ohio is requiring every public school to have a formal AI policy by July 2026.

Where This Lands

The impulse to mock a teacher is as old as school itself. What's new is that a 13-year-old can now produce a hyper-realistic explicit video of their teacher in minutes, post it to a platform where it gets 100,000 views overnight, and trigger consequences that follow that teacher for the rest of their career. Whether you think that's a fundamentally new problem or the same old problem with better tools probably determines where you land on punishment. But the surge in AI-generated abuse material suggests the tools aren't just amplifying the old behavior — they're creating a category of harm that didn't exist before. And right now, only 19% of students have heard a single word about it from their schools.

Sources