Policy

Federal Court Blocks Charge for AI-Generated CSAM Possession

Seventh Circuit rules existing Supreme Court precedent protects private possession of synthetic images, while production and distribution charges remain.

Omega Editorial· August 31, 2026· 4 min read

Court applies decades-old precedent to AI-generated images

A federal appeals court has dismissed a criminal charge against a Wisconsin man accused of privately possessing AI-generated sexual images of fictional children, finding that Supreme Court precedent from the analog era protects that narrow category of conduct under the First Amendment. The ruling leaves intact separate federal charges for producing, distributing, and sending such material to a minor.

The U.S. Court of Appeals for the Seventh Circuit issued its decision on August 25, nearly two years after investigators began examining Steven Anderegg's use of Stable Diffusion to allegedly create hundreds of sexually explicit images depicting prepubescent children. The investigation originated from a Meta report to the National Center for Missing and Exploited Children in October 2023, according to Law Commentary, which first reported the details.

The constitutional framework at issue

The case hinges on how courts distinguish between images involving real victims and entirely synthetic material. Federal prosecutors charged Anderegg under 18 U.S.C. Section 1466A, an obscenity statute that covers computer-generated images and explicitly does not require proof that a depicted child actually exists. Authorities acknowledged the images in question do not depict or identify any real child.

That distinction determines which constitutional rules apply. The Supreme Court established in Stanley v. Georgia (1969) that the government generally cannot punish someone for possessing obscene material in their home. The court later carved out an exception for child sexual abuse material involving real children, recognizing in decisions like New York v. Ferber and Osborne v. Ohio that actual victims are harmed both during production and through continued circulation of images.

Virtual images presented a different question. In Ashcroft v. Free Speech Coalition (2002), the Supreme Court struck down portions of a federal law that broadly prohibited sexually explicit depictions of fictional minors, finding that without an actual child victim, the government could not automatically apply the same rules.

Congress responded with the PROTECT Act in 2003, which included the obscenity provision now being used against Anderegg. The Seventh Circuit did not strike down that law entirely. Instead, it issued an "as applied" ruling, meaning the statute remains valid but cannot constitutionally be used to punish the specific conduct alleged in this case.

Prosecutors cite AI realism and investigative burden

Federal prosecutors argued that generative AI has fundamentally changed the landscape the Supreme Court considered two decades ago. Synthetic images can now appear so realistic that investigators may struggle to determine whether a real child has been abused, potentially diverting resources from identifying actual victims. The government also raised concerns about such images being used to groom minors.

The three-judge panel acknowledged these concerns. Judge John Z. Lee, joined by Judge Michael Kolar, noted that today's AI systems can produce images extremely difficult to distinguish from depictions involving actual children and called for additional Supreme Court guidance. However, lower courts must continue following existing precedent unless the justices change it.

Why it matters

This ruling exposes a growing tension between First Amendment doctrine developed in the pre-AI era and the capabilities of modern generative systems. While the decision protects a narrow form of private possession, it leaves prosecutors free to pursue charges for production, distribution, and transfer of such material—activities that extend beyond the home and potentially involve real harms like grooming. The case signals that courts may need updated guidance from the Supreme Court as AI-generated content becomes increasingly indistinguishable from reality, raising questions about how existing constitutional frameworks apply to synthetic media that can be weaponized even without direct victims.

Anderegg's remaining federal charges alleging production, distribution, and transfer of obscene material to a minor are pending in the U.S. District Court for the Western District of Wisconsin. The details of this case were first reported by Law Commentary.

#ai-generated content#child safety#first amendment#generative ai#stable diffusion#legal precedent

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 3 min read

Sony, Warner Sue Anthropic Over Alleged Music Copyright Theft

Major labels claim Claude AI was trained on thousands of copyrighted songs without permission, seeking damages up to $150,000 per work.

Via AI Watch · Aug 30, 2026
Policy· 2 min read

MIT Finds AI Can Complete Most Undergraduate Assignments

A new university report warns that generative AI models now produce credible solutions to essays, coding problems, and proofs across the curriculum.

Via AI Watch · Aug 30, 2026
Policy· 3 min read

Florida Proposes AI Use Rules for All State Colleges

The Department of Education wants each of the state's 28 college system institutions to establish policies governing artificial intelligence on campus by fall.

Via AI Watch · Aug 30, 2026