by Elizabeth Dorokhina, Program Intern
In an increasingly digital world, where 97% of US teens use the internet daily and 4 in 10 say they are almost constantly online, knowledge and prevention about online instances of sexual harm are becoming even more important (Pew Research Center, 2025). As teenagers spend more time online, there are also more instances of “deepfake nudes.” These are fake, intimate, nonconsensual images created using generative AI.
Thorn, a child safety non-profit focused on creating technology to defend children from sexual abuse and exploitation, released data showing 1 in 10 minors know of cases where friends have created “deepfake nudes” of other minors. Furthermore, one in 17 children are victims of ‘deepfakes’ (Thorn, 2025). This is at least one child in every classroom.
Though they are commonly called ‘deepfakes,’ the harm is anything but fake. Deepfake abuse uses AI-generated child sexual abuse to normalize the sexual exploitation of minors. This leads to nearly two-thirds of children worrying that AI could be used to create synthetic sexual images (UNICEF, 2026). The non-consensual distribution of intimate images (NDII) has even been called an epidemic (Internet Matters, 2024).
There are several known effective methods to stop NDII. Changes in criminal and civil laws are often considered the first line of defense¹, but promoting social norms that protect against violence are equally as important (Basile et al., 2016). Similarly, multiple studies have also shown that there is a correlation between rape-supportive attitudes and sexual aggression (Lonsway & Fitzgerald, 1994; Truman, Toaker, & Fischer, 1996). Standing against harmful social norms and rape-supportive attitudes can be a useful method to counter rising rates of sexual violence.
A study by the Journal of College Student Development explains that when men had a support group, they were more likely to challenge sexist environments (Barone & Wolgemuth, 2007). Peer conversations can act as an effective prevention tool for several reasons. Having supportive peers, as the study shows, can make men more likely to disrupt harmful conversations and actions. Also, having a community of peers that don’t have rape-supportive attitudes can shift social norms away from harmful language and behaviors, like NDII.
Though these conversations may seem difficult, preparing for them can include education, getting support, and making a plan (National Sexual Violence Resource Center). When engaging in the conversation with a peer creating deepfakes or AI-generated sexual abuse, it is important to listen in a nonjudgmental manner without approving their choices (Stanford). Being educated before beginning the discussion ensures that the individual is able to communicate with their friend about boundaries, consent and the potential harm of their actions. Furthermore, if talking to a student, knowing relevant university or school policy can help indicate to the peer the consequences of their actions. Also, it is important to practice everyday interventions when smaller issues arise so it becomes easier to address larger concerns (Russo, 2013). Peer conversations may be difficult, but they don't have to be. The National Sexual Violence Resource Center (NSVRC) has a guide for young people on how to initiate and hold these types of conversations.
Another study shows how peers tend to influence others. In this study, peers that perpetrate sexual violence were more likely to have friends that also perpetuate sexual violence (Esperage et al., 2021). Having friends that perpetuate sexual violence also made an individual more likely to do the same (Casey & Lindhorst, 2009). Therefore, when looking at issues of sexual harm created through generative AI, it is also important to look at peer relationships. This can help stop the spread of negative influence. Peer relationships and accountability within these relationships is especially important for preventing NDII and to prevent actions associated with NDII becoming acceptable behaviors.
When aiming to prevent harm, peer to peer conversations are an effective strategy for sexual assault that happens through the use of technology or otherwise (Casey & Lindhorst, 2009). With deepfakes, in particular, having conversations about consent, serious consequences of making a deepfake, and victim experience can all be helpful tools for combatting nonconsensual distribution of intimate images (Amaze, 2026). By encouraging conversation about rape-supporting attitudes, limiting peer pressure, and active bystander intervention, teens can limit risk factors for friends to engage in acts of sexual violence
1. MCASA advocated for SB360, a bill that was just enacted in 2025 and addresses deep fakes.
References
Amaze. (2026). “Deepfakes/AI Nudes.” Amaze. Retrieved here.
Barone, R.P., Wolgemuth, J.R., & Linder, C. (2007). Preventing Sexual Assault Through Engaging College Men. Journal of College Student Development 48(5), 585-594. Retreived here.
Basile, K.C., DeGue, S., Jones, K., Freire, K., Dills, J., Smith, S.G., Raiford, J.L. (2016). Sexual Violence Prevention Resource for Action: A Compilation of the Best Available Evidence. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. Retrieved here.
Casey, E. A., & Lindhorst, T. P. (2009). Toward A Multi-Level, Ecological Approach To The Primary Prevention Of Sexual Assault: Prevention in Peer and Community Contexts. Trauma, Violence & Abuse, 10(2), 91–114. Retrieved here.
Espelage, D.L., Rulison, K.L., Ingram, K.M. et al. (2022). Social Networks of Adolescent Sexual Violence Perpetrators: Peer Friendship and Trusted Adult Characteristics. Prev Sci 23, 154–166 Retrieved here.
Faverio, M. & Olivia Sidoti. (2025). “Teens, Social Media and AI Chatbots 2025.” Pew Research Center. Retrieved here.
Internet Matters Team. (2024). “New report estimates half a million UK teenagers have encountered AI-generated nude deepfakes.” Internet Matters. Retrieved here.
National Sexual Violence Resource Center (NSVRC). (n.d.). Let’s Talk: How to Have the Conversation. NSVRC. Retrieved here.
Ringrose, J., Horeck, T., & Rodda, E. (2026). Sexualized Deepfakes in UK Schools: Understanding and Preventing AI-Generated Image-Based Sexual Abuse Through Better AI Literacies. Behavioral sciences (Basel, Switzerland), 16(4), 554. Retrieved here.
Russo, A. (2013, September 13). 10 Strategies for Cultivating Community Accountability. Prison Culture. Retrieved here.
Stanford SHARE Title IX and Title VI Office. (n.d.). Supporting a Friend Who's Caused Harm. Stanford University. Retrieved here.
Thorn. (2024). Deepfake Nudes & Young People: Navigating a new frontier in technology-facilitated nonconsensual sexual abuse and exploitation. Retrieved here.
UNICEF. (2026). “‘Deepfake abuse is abuse’.” UNICEF. Retrieved here.
“10/8/25: MCASA's VCT #2: Emerging Issues in Digital Safety & Sexual Violence on Campus.” (2025). MCASA. Retrieved here.
