A federal appeals court has ruled that the First Amendment protects the in-home possession of artificial intelligence-generated child sexual abuse material, a decision that stems from existing Supreme Court precedent. The Seventh Circuit stated that it was bound by a 2002 Supreme Court decision, Ashcroft v. Free Speech Coalition, which limited restrictions on sexually explicit depictions of fictional children when no actual child was involved in their creation.
Judges on the Seventh Circuit expressed concern that the rapid advancement of AI technology has outpaced legal frameworks established nearly 25 years ago. The court affirmed a lower court's dismissal of a charge against Steven Anderegg, who possessed hundreds of AI-generated sexually explicit images depicting children on devices seized from his home. Anderegg faced charges including production, distribution, and possession of the material.
The Justice Department had appealed the dismissal of the possession charge, arguing it should be able to prohibit the possession of virtual child sexual abuse material even without a real child's involvement. However, the appeals court sided with the lower court, adhering to the principle that the government generally cannot criminalize the private possession of obscene material within a home. This distinction is complicated by the fact that child sexual abuse material involving real children is prohibited due to the exploitation of actual victims.
In the 2002 Free Speech Coalition ruling, the Supreme Court reasoned that the government's interest in protecting children did not apply with the same force when no real child was depicted or harmed. However, the Seventh Circuit judges noted that the AI image-generation technology available today was likely unimaginable at the time of that ruling. They warned that AI can now generate images that are virtually indistinguishable from material depicting the abuse of real children.
One of the judges, John Lee, wrote, "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." Another judge, Joshua Kolar, concurred, suggesting that the Supreme Court may need to revisit its precedent due to the proliferation of AI technology.
Concerns have also been raised that the rise of AI-generated child sexual abuse material could impede investigators trying to identify real victims, as increasingly realistic artificial images become harder to distinguish from depictions of actual abuse. Judges Lee and Kolar also cited research indicating that some AI training datasets may contain hundreds of child sexual abuse material images, potentially creating a link to the exploitation of real children, though they stressed this issue was not formally raised or sufficiently explored in this specific case.
Justice Clarence Thomas, in a concurring opinion in the 2002 Supreme Court ruling, anticipated this dilemma. He warned that if technological advances hinder the prosecution of unlawful speech, the government might have a compelling interest in regulating certain categories of lawful speech to enforce laws against material produced through the abuse of real children. Meanwhile, First Lady Melania Trump has achieved a victory under her Take It Down Act with the first conviction in a case involving cyberstalking and the use of AI to create obscene images of adults and children, though this ruling addresses a different aspect of AI misuse.