When a student creates an AI deepfake of a classmate, they almost always do it from home, on a personal device, outside school hours. In 2021, the Supreme Court decided Mahanoy Area School District v. B.L., the case of a Pennsylvania cheerleader who got suspended for posting a profane Snapchat after she didn't make varsity. The Court ruled 8 to 1 that the school couldn't discipline her for off-campus speech. The decision has been read broadly as protecting students from school punishment for things they post on personal devices on their own time.

But Mahanoy did not draw the line where some districts have decided it did. Justice Breyer's majority opinion was explicit about what schools can still regulate even when the speech happens off campus. His words: "serious or severe bullying or harassment targeting particular individuals."

That carveout is the doorway. AI deepfakes of identifiable students, distributed to other students, fit cleanly through it. The mistake some districts are making is reading Mahanoy as a hands-off rule. It is not. It is a heightened-scrutiny rule. Schools that document the disruption and target their response to the specific harassment have a defensible legal position.

The technology is new. The legal category is not.

AI-generated nude images of students that circulate among classmates and contribute to a hostile educational environment are, by any reasonable reading, sex-based harassment. Schools have had Title IX obligations to investigate and remediate this category for decades. The fact that the images were AI-generated does not relieve a district of those obligations. Harassment continues for as long as the file exists on someone's phone.

On May 19, 2025, the TAKE IT DOWN Act was signed into law. It originated after a 2023 incident at Aledo High School in Texas, where a student created and shared photo-manipulated images of nine female classmates and the local school district found itself unable to act.

The Act does two things. First, it makes the knowing publication of non-consensual intimate imagery, including AI-generated deepfakes of identifiable people, a federal crime. Second, it requires online platforms to set up takedown processes that remove flagged content within 48 hours of a valid request.

Will every platform actually comply? Probably not. Will the students and families who need this remedy know it exists? Probably not, unless someone tells them. That is part of the school's role now.

If you are leading a school or district right now, three things matter.

1. Treat any AI-generated intimate imagery of a student as a Title IX incident from the first report. Document it. Provide supportive measures to the targeted student. Investigate. Do not put the targeted student back into proximity with the alleged creator before the investigation is complete.

2. Re-read your acceptable use policy. Most district AUPs were written before generative AI was widely available. The phrase "computer-generated nude images of classmates" almost certainly does not appear anywhere in your policy. It should.

3. Train your front-line staff to recognize this as a category of harm. Counselors, teachers, bus drivers, club sponsors. The students who experience this are not always going to use the words "AI deepfake." They are going to say someone made a picture of them. The adults around them need to hear that and act on it.

The harder question, and the one I do not see administrators talking about as openly, is what to do about the kids who do this. Prevention beats both cure and punishment.

Most of them are not strangers on the internet. They are eighth graders with a smartphone and an app that did not exist three years ago. They are kids who do not fully understand what they did, even when what they did was a federal crime as of last May. They need consequences. They also need education that almost none of them are currently receiving in any school.

That is the work I do at Beyond the Bell Co. The first lesson in the Screen Smart unit, Real or Robot?, is free and gives middle and high school students a baseline for thinking about AI-generated content. It does not solve the deepfake problem on its own. It is one piece of a much larger conversation that needs to be happening, district by district, this school year.

If you are in a district that has not had this conversation yet, the conversation is overdue.