Dr. Mary Anne Franks gets why people may be upset by a recent court ruling on possessing AI-generated child sexual abuse material.
The decision, issued in late August by a federal appeals court panel, found that the First Amendment protects the possession of child sexual abuse material, or CSAM, in the home if it was created using artificial intelligence and does not depict an actual child.
“A lot of people have quite understandable intuitions about how wrong this sounds, or how it sounds like there’s something radical going on here,” said Franks, the Eugene L. and Barbara A. Bernard Professor in Intellectual Property, Technology, and Civil Rights at the George Washington University Law School.
Instead, Franks said the panel adhered to Supreme Court decisions issued decades ago, before AI could render hyper-realistic images of fake people and fake images of real people.
“I do think it’s a really straightforward reading of prior precedent, but I also understand why it’s really upsetting to hear,” Franks told Mashable.
But that’s not the end of the story. The judges made clear their discomfort with the conclusion by signaling to the Supreme Court that it might reconsider the precedent. Franks called the move “unusual.”
“[G]iven the significant advances in computer-generated image technology, particularly in the field of artificial intelligence, we would benefit from additional guidance from the Supreme Court regarding the intersection between the First Amendment and virtual CSAM if an appropriate case should arise,” wrote Judges John Z. Lee and Joshua P. Kolar.
What is the AI CSAM case really about?
The three-judge panel, which is part of the U.S. Court of Appeals for the 7th Circuit, considered the facts of a federal case against Steven Anderegg, a software engineer who used generative AI tools to create “hyper-realistic” CSAM.
Not only did he allegedly produce, distribute, and possess the material, but he also sent it to a minor via Instagram direct messages in Oct. 2023. Meta, Instagram’s parent company, detected and reported the image and reported it to the National Center for Missing and Exploited Children’s CyberTipline.
A law enforcement investigation subsequently led to Anderegg’s arrest. He was charged with multiple crimes, including knowingly possessing “at least one visual depiction that depicted a minor engaging in sexually explicit conduct and was obscene.”
Anderegg moved to dismiss the possession charge, arguing that conviction for possession of CSAM would violate his First Amendment rights as defined by past Supreme Court decisions. A district court agreed. Prosecutors appealed that decision, landing the case in the U.S. Court of Appeals for the 7th Circuit, which ultimately dropped the possession charge.
Anderegg still faces charges for producing and distributing CSAM.
Why is this kind of CSAM protected by the First Amendment?
A combination of past Supreme Court cases led to the dropping of Anderegg’s CSAM possession charge, Franks explained.
Namely, in 1969, the Supreme Court ruled in Stanley v. Georgia that individuals have a right to possess obscenity in their own home.
By 1990, however, the Supreme Court carved out an exception when it addressed the possession of “child pornography” in Osborne v. Ohio. The decision recognized that, possession of CSAM, even if viewed only in the home, serves both as a permanent record of a child’s abuse and as a potential lure for other victims. Prohibiting possession of this material could help decrease the demand and production of CSAM, the justices reasoned.
Then, in 2002, the Supreme Court considered a federal law provision that prohibited the possession of virtual CSAM. The ban was broadly written to include any visual depiction or computer-generated image that appeared to be of a minor engaging in sexually explicit conduct.
The court struck down the provision and effectively declared that entirely computer-generated CSAM is protected by the First Amendment.
The Court did not address another provision of the federal law that banned “morphed” CSAM, or computer-generated CSAM created using images of real children.
The U.S. Court of Appeals for the 7th Circuit judges who recently relied on the 2002 ruling to decide Anderegg’s appeal acknowledged the difficulty of doing so.
“Given the relentless advancement in artificial intelligence models, we have some concerns about the lines these cases draw, but we are not free to redraw them ourselves,” Judge John Z. Lee wrote in his opinion.
What happens next with AI CSAM
The government has not yet appealed the 7th Circuit’s decision, but Franks said it has the option to do so. In that scenario, the Supreme Court would consider taking up the case in the October term.
Franks said the conclusion that fully virtual CSAM is protected by the First Amendment could be successfully challenged. In particular, the government could argue that AI models have indeed been trained on CSAM imagery.
As a result, even if a particular AI-generated CSAM image doesn’t depict a real child, it may have been produced using an image of a real child being victimized.
Franks does believe that prohibitions of virtual CSAM must be very narrowly and carefully defined to avoid “weaponization by moralistic crusaders with anti-LGBTQ+, misogynist, and/or anti-sexual education agendas.”
The Heritage Foundation’s Project 2025, for example, argues that porn should be outlawed and those who produce and distribute it should be imprisoned. Project 2025 links porn to “the omnipresent propagation of transgender ideology and sexualization of children.”
Still, Franks said the risks of AI CSAM are real and present, not hypothetical possibilities.
In her opinion, AI CSAM normalizes the sexualization of children by adults and by other children, undermining their bodily autonomy, self-esteem, and flourishing.
“The question we should be asking is what possible benefit AI CSAM provides that justifies such harms,” Franks said.
If you are a child being sexually exploited online, or you know a child who is being sexually exploited online, or you witnessed exploitation of a child occur online, you can report it to the CyberTipline, which is operated by the National Center for Missing Exploited & Children.