On March 25, two 16-year-old boys from Lancaster, Pennsylvania received probation for using generative artificial intelligence to create fake child sexual abuse material(“CSAM”) of their female classmates. Both defendants admitted in court to participating in the criminal scheme, including 59 separate felony counts of manufacturing child sexual abuse material. In total, the defendants created 347 photos and videos depicting 60 victims, all but one of whom were under the age of 18. They now must complete 60 hours of community service, may not have contact with any of the victims, and must pay $12,000 in restitution to cover counseling costs for their victims.
In the juvenile proceeding, prosecutors alleged the defendants, both 14 years old at the time of the offenses, used generative artificial intelligence to “morph” nude versions of photographs taken from social media, school yearbooks, and FaceTime video chats of 48 female classmates at Lancaster Country Day School and 12 other victims. The defendants shared the CSAM, which depicted their classmates nude or engaged in sexual activities, with each other on Discord, an online messaging service. In November 2023, one defendant inadvertently shared the CSAM with other students at the school.
Shortly afterwards, a tip was submitted on Safe2Say, Pennsylvania’s anonymous school safety tip service, notifying school officials about the CSAM. Lancaster Country Day School received the tip but did not contact police. As such, the delinquents continued creating the CSAM until May 2024. The criminal investigation only began when parents alerted law enforcement directly.
At the disposition hearing on March 25, victims spoke about the harm the defendants’ actions had caused them. One victim told Judge Brown, “I will never understand why they did this,” saying the CSAM “destroyed my innocence.” Another victim criticized one of the defendants for expressing “fake empathy” to female classmates who confided in the defendant about the hurtful images before the defendant’s participation in their creation was made known. Judge Brown stated at the disposition hearing that he had not heard the defendants apologize or take responsibility for their actions.
This case reflects a disturbing national trend of AI-generated CSAM. Beginning in mid-2023, cases emerged of male middle school and high school students using AI tools to “nudify” images of real women and girls to create “deepfake” pornography and CSAM. Several states, including Pennsylvania, have enacted legislation targeted at AI-generated CSAM. In October 2024, Governor Josh Shapiro signed Act 125 into law, which criminalized sharing AI-generated sexual depictions of real people with the intent to “harass, annoy or alarm another person.” Disseminating AI-generated sexual images is a first-degree misdemeanor under PA law, punishable by up to 10 years in prison and a $10,000 fine, if the victim is a minor. At the federal level, the President signed the Take it Down Act into law in May 2025, which criminalizes the online sharing of AI-generated sexual images of real people and requires certain covered platforms to promptly remove such images upon receiving notice.
Deepfake pornography, which is generated by feeding non-explicit images into AI without the subject’s permission, is a pernicious tool of gender oppression: A 2023 report indicated that deepfake pornography accounted for an astonishing 98% of deepfake content on the Internet, and that 99% of deepfake pornographic images and videos depicted women and girls. Offenders, largely male, use nonconsensual, AI-generated pornographic images to harass and abuse real women and girls. The psychological effects on victims are profound: more than half of deepfake victims have contemplated suicide. Young women and girls are particularly vulnerable to being targeted and harmed by the creation of such images.
While we commend Lancaster County prosecutors for diligently pursuing this case, and federal and state authorities for passing new laws to address the scourge of AI-generated CSAM and pornography, we believe existing efforts do not go far enough. Pennsylvania’s Act 125, along with the Take it Down Act, target the dissemination—rather than the production or possession—of AI-generated CSAM. This framework ignores the fact that the creation of such images, even without publication, causes harm by fueling sexual interest in children, normalizing extreme violence, and increasing the risk of contact offending. Lawmakers must get serious about the harm caused by the creation of AI-generated CSAM and impose criminal penalties to deter the creation of such content.
All views expressed herein are personal to the author and do not necessarily reflect the views of Villanova University Charles Widger School of Law or of Villanova University.