Seventh Circuit Court of Appeals says it was bound by decades-old Supreme Court precedent in ruling on AI-generated child sexual abuse material
A Wisconsin criminal case has put a disturbing question before our nation’s courts — what happens when technology advances faster than our nation’s laws and constitutional precents designed to regulate it?
That question was raised this week after the U.S. Court of Appeals for the Seventh Circuit in Chicago upheld the dismissal of a federal possession charge against a Wisconsin man who was accused of creating thousands of sexually explicit AI-generated images depicting children.
Three other charges are pending against Steven Anderegg, 42, of Holmen, Wisconsin. The court’s decision determined that under existing Supreme Court precedent, the government cannot criminalize Anderegg’s private in-home possession of AI-generated images that do not depict actual children.
Even the judges who issued the ruling acknowledged the outcome is unsettling, but their hands were tied as a result of decades-old Supreme Court precedents
According to federal prosecutors, Anderegg used a free AI-generative software program called Stable Diffusion to create realistic images depicting prepubescent children engaged in sexually explicit conduct.
The government also alleged that similar AI-generated material was sent to a minor. Anderegg was charged in federal court with producing, distributing and possessing visual depictions of minors engaged in sexually explicit conduct, along with transferring such material to a child under the age of 16.
The case began after Meta, the parent company of Facebook and Instagram, reported an account to the National Center for Missing and Exploited Children. Investigators subsequently searched Anderegg’s home and electronic devices.
In 2025, U.S. District Judge James Peterson dismissed the possession charge but allowed the other charges to proceed.
The Department of Justice appealed the possession ruling, arguing that federal law should permit the government to prohibit possession of obscene virtual child sexual abuse material even when no actual child was involved in producing the images. The Seventh Circuit Court disagreed.
The court says its hands were tied
August 25, 2026, the decision issued was centered on two Supreme Court precedents that predated modern generative AI technology by decades.
In the Stanley v. Georgia case in 1969, the Supreme Court held that the government generally cannot criminalize an individual’s private possession of obscene material inside the home.
Then, in Ashcroft v. Free Speech Coalition in 2002, the Supreme Court struck down federal restrictions on certain sexually explicit virtual images involving fictional minors when no actual child was involved in their creation.
The Seventh Circuit judges noted those decisions determined the outcome in the Anderegg’s possession charge. In other words, the appeals court wasn’t ruling that the material was harmless or that AI-generated sexual images of children should be illegal. The ruling says a lower federal court cannot disregard Supreme Court precedent because the technology has changed.
Judge John Lee was joined by Judge Joshua Kolar in a concurrence to make the unsettling ruling particularly clear — until the laws change, the courts hands are tied.
“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,” Lee wrote.
That may be the most important sentence in the entire ruling.
Technology has changed, but the precedent hasn’t
When the Supreme Court decided Ashcroft v. Free Speech Coalition in 2002, modern generative AI systems were nothing more than science fiction out of the 1999 Matrix film staring Keanu Reeves. Today’s technology is capable of producing photorealistic images from a simple text prompt within seconds.
These realistic images are difficult to distinguish from actual photographs, particularly to the human eye. The Seventh Circuit judges specifically acknowledged that technological advancements have complicated the legal distinction between fictional and real child abuse imagery.
The judges also raised concerns about research indicating child sexual abuse material (CSAM) has appeared in various datasets used to train some AI image-generating systems, potentially creating a connection between synthetic imagery and the exploitation of actual children.
The judges noted the government did not adequately raise that issue in this case, and the existing record was insufficient for the court to decide it. That leaves our judicial system in a troubling technological paradox.
The law distinguishes between an image depicting a real child and a computer generated image. However, AI increasingly makes that distinction harder for humans to see.
The Wisconsin man still faces three charges
Headlines surrounding the case easily create the impression that Anderegg was cleared, but that is not exactly what happened. The Seventh Circuit dismissed one charge — the private possession of the AI-generated material.
The remaining allegations involving production, distribution and transferring material to a minor remains in place. The ruling does not establish a constitutional right to create or distribute AI-generated sexual images of children. Instead, it addresses whether the federal government can prosecute someone for possessing entire virtual material inside his home when no actual child was involved. Under current Supreme Court precedent, the answer was no.
Wisconsin lawmakers already confronting the AI problem
The Anderegg case is particularly significant in Wisconsin because state lawmakers have already begun confronting the same technological / AI problem.
In March 2024, Wisconsin enacted a law specifically addressing AI-generated sexually explicit images depicting children. A separate case in Eau Claire County has already tested that law’s constitutionality.
In that case, prosecutors charged a 24-year-old Altoona man after allegedly discovering AI-generated images depicting babies engaged in graphic sexual acts. His attorney challenged the constitutionality of Wisconsin’s law. Kai Ahmed Ender Jon was charged with six felony counts of possession of virtual CSAM. Each count carries a maximum penalty of 15 years in prison and up to $100,000 in fines.
The Supreme Court will have to revisit the issue
The Seventh Circuit’s decision could pressure the Supreme Court to revisit the legal framework surrounding virtual child sexual abuse material.
Justice Clarence Thomas anticipated part of this problem in his 2002 Free Speech Coalition dissent, warning that technological advances could eventually make virtual depictions indistinguishable from images involving real children.
More than two decades later, that hypothetical concern is becoming reality. The Seventh Circuit judges recognize the problem, but they also recognize their limits.
That leaves Congress and the Supreme Court with an urgent problem. The laws that were written before Artificial Intelligence existed will need to be rewritten to address technology capable of creating synthetic images of children and CSAM.
The real question is whether lawmakers are capable of acting swiftly, and whether they are capable of drawing that metaphorical demarcation line without violating our constitutional protections for free speech and expression.
Wisconsin finds itself at the center of a national debate — if AI can now create child abuse images, is the law truly equipped to protect children? A Supreme Court case is likely around the corner. The Heartland Post will continue to follow this story.
