AI explicit deepfake apps reaching classrooms, student-led AI policy, and the privacy cost of AI assistants
Key Takeaways
- AI “nudify” apps remain within students’ reach. Researchers posing as 13-year-olds downloaded 135 apps that make fake explicit images, despite app store bans and parental controls.
- Students are helping write school AI policy. Ninety-eight students from all 50 states wrote the STUDENT FIRST Act, and Day of AI’s new Students First Tour is helping them bring it to their districts and states.
- AI assistants can act on your behalf, raising the privacy stakes. Meta’s Muse and OpenAI’s new agents take actions for users by tapping personal data, and in one case Muse shared a user’s home address with a stranger.
AI “nudify” apps are banned from the app stores, yet students can still download them
Despite outright bans, teens can still download AI apps that make fake explicit images, and the harm is reaching classrooms.
The Gist
Apple and Google both ban AI “nudify” apps, the tools used to make fake explicit images of real people, from their app stores entirely. Even so, a new investigation found researchers posing as 13-year-olds could still download 135 of them, and the apps got past parental controls too. The harm is already reaching classrooms, where the National Center for Missing & Exploited Children (NCMEC) says a growing share of these cases involve students creating deepfakes of classmates. App-store bans can only do so much, so schools play a key role in educating students on the impact of explicit deepfakes and how to respond when they encounter one.
What to Know
The investigation, from the Tech Transparency Project, tested how easy these apps are to reach. Researchers set up accounts as 13-year-olds and downloaded 135 apps that could digitally undress people in photos or paste their faces onto explicit images. Apple and Google both ban these apps, but they rely on developers to describe and rate their own apps. The companies only remove them once they’re reported. That’s how these apps made their way into the stores despite the ban, and because developers also rated many as safe for minors, even accounts with parental controls could download them.
The problem is growing quickly. NCMEC reported that CyberTipline reports involving AI-generated sexual images jumped from 4,700 in 2023 to 440,000 in the first six months of 2025. A Common Sense Media survey of teens found that 44% had seen AI-generated sexual material, and nearly a quarter of them said it showed someone they knew. Common Sense Media Director of Research Supreet Mann described the moment as the “Wild West,” adding that the tools are so easy to use that a student “can take any image of any young person and create this.”
This is where educators can make a real difference. Many students don’t realize how serious this content is, or that creating an explicit image of a minor is against the law, and only 13% say they’d feel comfortable telling a teacher or counselor. When educators understand the issue, they can be the trusted adult who helps students recognize the harm, respond the right way, and feel safe coming forward.
TSI’s Take
The investigation shows that bans and parental controls alone can’t keep these apps out of students’ hands. Schools can shape how students respond when a fake appears, and whether they choose to create one in the first place. You can prepare students by:
- Naming the harm clearly: Teach students that making an explicit image of a minor is illegal and seriously hurts a real person.
- Giving them a response plan: Make sure students know to stop, save evidence, tell a trusted adult, and report it if a fake appears.
- Building a safe way to report: Create a clear, low-stakes path to tell an adult, since most students say they wouldn’t feel comfortable bringing it up.
Give students a clear plan for the moment a deepfake appears with the #WinAtSocial Lesson, Explicit AI deepfakes: ethics, laws, and response options, where students practice responding and protecting one another.
Students helped write a national AI policy, and now they’re bringing it home to their own schools
Students from all 50 states wrote a model AI policy, and a new tour is helping them bring it to their own districts.
The Gist
A new movement is giving students a real hand in shaping their schools’ AI policies. Over the summer, 98 students representing all 50 states wrote the STUDENT FIRST Act, a framework for how AI should be used in K-12 schools. Now the nonprofit behind it, Day of AI, is launching the Students First Tour to help those students bring the ideas home, leading AI policy discussions in their own districts and states. The point educators and students keep landing on is that the people using these tools every day should help set the rules, and schools that bring student voice into their policies end up with guidelines students are more likely to understand and follow.
What to Know
According to GovTech, the STUDENT FIRST Act came out of the Student Senate at America’s Youth AI Festival in July, where 98 students from every state drafted and approved a shared framework for responsible AI use. The new Students First Tour, run by Day of AI, keeps that work going with three parts: a national student ambassador cohort that will lead local AI policy discussions, ready-made guides so any school can run its own student senate, and an online repository where students can submit their own policy ideas.
The effort has real reach. Day of AI is working with the Edward M. Kennedy Institute for the U.S. Senate, MIT RAISE, and the School Superintendents Association, which is sharing the framework with more than 10,000 school leaders. Students aren’t waiting for permission to use it, either. In Oregon, they presented the Act to the state Senate’s education committee, and in Connecticut, students brought it to their State Board of Education.
That momentum matches a bigger shift in the field. The opening keynote at the 2027 Future of Education Technology Conference, District Administration reports, is titled “Power Players: The Next Generation Is Already in the Game,” and it puts a high school senior, Yash Maheshwari, on stage to urge educators to keep student voice at the center of AI decisions. As conference chair Jenn Womble put it, some of the most powerful voices in education “are not waiting for permission—they are already creating the future of learning.” The takeaway for educators is that bringing students into AI policy makes the rules more credible and more workable for the people who live by them.
TSI’s Take
When students help write the rules, they build policies that hold up in real classrooms, and they practice using their mic for good to create positive change in their community. Schools don’t need a national tour to start, just a real seat for students at the table. You can give students that voice by:
- Including them in decision-making: Have students share their thoughts when your school writes or revisits its AI guidelines.
- Getting their insights: Survey students on where AI helps their learning and where it gets in the way before setting rules.
- Giving their ideas a real audience: Let students present their proposals to a principal, the school board, or student council.
Help students shape the tech rules they live by with the #WinAtSocial Lesson, Student voice in school tech and AI policies, where students think through what fair, workable guidelines look like.
AI assistants can now act on students’ behalf. What does that mean for their personal data and privacy?
New AI agents can take actions for you, which makes knowing what personal information to share a must-have skill.
The Gist
A new kind of AI is here, and it does more than answer questions. AI agents like Meta’s Muse and OpenAI’s new assistants can take actions for you, booking reservations, managing messages, even making deals. One user, according to the Associated Press, let Muse handle a Facebook Marketplace sale, and it shared his home address with a stranger who showed up at his house at night. These tools work by collecting a lot of personal information, so the more they can do, the more they know about you. As students start navigating assistants like these, knowing what information to share and withhold from AI tools becomes a critical skill for their safety and well-being.
What to Know
According to the Associated Press, AI agents are a step beyond chatbots. Instead of just answering a prompt, they run in the background and take actions on your behalf. When content creator Matt Robb let Meta’s Muse manage a Marketplace listing, the agent struck a deal on its own and gave out his home address, and a buyer he didn’t know turned up outside his house late at night. It came down to the permissions he had granted when he connected Muse to his account.
These assistants are only as useful as the information they can reach, so the more they know about you, and sometimes about your friends and family, the more actions they can take. Assistant professor at Georgia Tech’s School of Cybersecurity and Privacy, Veronica Rivera, put it this way: “Using an AI agent is similar to handing someone your keys and a to-do list. How helpful they are, and by extension how risky, depends on which keys you hand over.”
While this story isn’t set in a classroom, it still lands on a skill students need. The AI tools they use may request access to their accounts, contacts, and personal details. Knowing what to share, and what to hold back, is the difference between a helpful assistant and a privacy risk.
TSI’s Take
The lesson for students is the same one that can sometimes trip up adults: convenience can cost you privacy when you hand over more than you mean to. Schools can help students protect their privacy like they’re famous by building the habit of checking what an app can see before they say yes. You can build that habit by:
- Reading before allowing: Have students pause on permission pop-ups and ask what an app really needs.
- Sharing the minimum: Teach students to give an app only the access the task requires.
- Thinking past themselves: Remind students that their contacts’ and friends’ information can be exposed too.
Help students take control of what they share online with the #WinAtSocial Lesson, AI data privacy, where students practice identifying what information they would and wouldn’t share with AI tools.
#WinAtSocial gives K-12 educators ready-to-teach lessons on the skills in these stories, from responding to AI deepfakes to shaping school tech policy and protecting personal privacy, built around what students actually face online. Request a demo to see how it works in your school.