Schools across Illinois have a new item to check off their list as they reopen: informing and educating students and parents about the consequences of using artificial intelligence to create deepfake videos, images, or audio content without consent.
Last July, the state passed a bill amending its regulations governing schools to include the generation and distribution of AI deepfakes as a form of cyberbullying. By July 1, schools were expected to update their cyberbullying policies to reflect this inclusion, pursue appropriate disciplinary measures, and educate students about AI misuse.
There has been a “Cambrian explosion” of AI misuse in schools, especially among students who have used free apps online to create sexually explicit videos or images of classmates, mostly female, without their consent. In the absence of a clear policy on how to deal with deepfakes, and with few preventative measures in place, schools have struggled to limit the psychological and emotional fallout for targeted students, or to establish strict consequences for perpetrators.
The new legislation is a step in the right direction because it gives schools an existing framework to apply to an emergent problem, experts said.
“It will help to clarify any confusion and potential arguments that [cyberbullying laws] would not apply,” said Daniel Krause, the principal of Willowbrook High School in Villa Park, Ill.
Krause has updated his school’s cyberbullying policies to include terms like artificial intelligence, deepfakes, and likeness, an exercise he started with other school leaders in his district even before the legislation passed.
“The legislation can help … students and families understand the seriousness around the improper [AI] use, so that we can really focus on the potential benefits and usages in education,” Krause added.
However, updated cyberbullying policies need to go beyond outlining consequences, said Kristin Woelfel, a policy counsel for the Center for Democracy and Technology, a nonprofit that advocates for online civil liberties. They also must establish protocols to support those impacted by deepfakes.
“It’s not just like ‘You’ll get in trouble if you do this.’ It should also be like, ‘Here is who you report something like this to.’ Even the initial reaction when a student tries to report something can be very consequential to how they like move past it,” said Woelfel.
Locating the new legislation in school policy
AI generated deepfakes and social media can be a dangerous combination, said Arwen Lyp, the principal of Downers Grove South High School in Downers Grove, Ill.
Before social media or easily accessible AI generators, bullying a student would last a few moments or days at best, said Lyp. But now, videos or images that go viral bully students repeatedly. While there is no confirmed number of such cases in Illinois or other states, a 2024 survey by the Center for Democracy and Technology found that 40% of K-12 public school students were aware of a deepfake involving someone at their school, including 15% who knew of a sexually explicit deepfake depicting a member of their school community.
AI deepfake cases often come bundled with other forms of cyberbullying, said Krause. Adding broad terminology like “inappropriate use of technology, utilizing the likeness of someone, or violating individual rights” will allow school leaders to identify and investigate varied offenses under the updated cyberbullying policies.
While Krause has added specific words to define AI-generated deepfakes in the new policy, Lyp’s school has left its wording unchanged. “The new legislation is encompassed in our policy by our word choice of cyberbullying,” she added.
While adding this to the existing cyberbullying policy makes sense, it could just as easily fit into another policy, said Woelfel. She outlined a few areas to focus on, based on CDT’s own model policy on nonconsensual intimate imagery:
- Identifying what a deepfake is, and how it’s created;
- Outlining consequences for students and families. Students should be aware that this behavior might lead to a felony charge;
- Defining reporting protocols, which include identifying adults in the school to report to;
- Providing support for students or teachers who are victimized by these deepfakes;
- Making public resources—such as support networks, reporting tools, and websites that can take down images—easily accessible.
Principals focus on prevention through education
No policy or procedure can fully repair the serious harm caused by an explicit deepfake image or video, said Krause. School leaders must address the problem even before it occurs.
Krause’s school year usually starts with grade-level meetings on student safety, focusing on how to report something that’s causing harm to them or their peers, or putting them in danger.
There’s a similar session for teachers to set the right expectations—if they witness cyberbullying among their students, or if a student brings them an inappropriate image or video, teachers should know whom to report to, without looking at the material themselves. Over the last few years, Krause has also included discussions about the improper use of AI in his meetings with parents and made it a part of his back-to-school message.
Other principals have expanded their digital citizenship lessons to include AI misuse and its impact. Jeff Wardle, the principal of Buffalo Grove High School in Buffalo Grove, Ill., has taken this a step further and linked online behavior to his school’s core values like kindness, bravery, and belonging.
“We want students to be curious about how AI works and its positives, but also how it can be manipulated for bad. And then we want students to be brave, … the see something, say something kind of mentality to speak up,” said Wardle. In the upcoming school year, Wardle plans to host an open training for all students on deepfakes, what they mean, their impact, and consequences.
He’s also planning for the school resource officer to address deepfakes when they talk to students about other topics like underage drinking, smoking, or creating pornography. The SRO’s involvement may help students understand the long-term legal consequences of creating and distributing deepfakes. “It’s not a scare tactic. It’s more about an education tactic,” he said.
Wardle wants to expand that education to parents as well. In his freshman orientation meetings with parents this year, Wardle will discuss how parents can help schools prevent such misuse of technology by sharing the names of common AI apps, how they can emphasize the need to be responsible digital citizens, and explaining how to report concerns to the school administration.
“We want them to partner with us. We can’t do it by ourselves, and frankly, they probably can’t do it by themselves. But if we partner together on this, I really feel like we can get ahead of it and hopefully stop it before it even starts,” said Wardle.
