After Anderegg: What the Seventh Circuit’s New Ruling Means (and Doesn’t Mean) for Indiana’s Virtual CSAM Statute
Last week the Seventh Circuit handed down an opinion in U.S. v. Anderegg, affirming the dismissal of a federal court charging the in-home possession of AI-generated CSAM. The court’s holding could best be described as narrow, reluctant, and explicitly bound by the U.S. Supreme Court’s precedent. In other words, it seemed clear the panel of judges did not want to rule the way they did, but they were bound to do so. The Indiana Court of Appeals issued a nearly identical opinion in 2024 in F.G. v. State, 244 N.E.3d 972 (Ind. Ct. App. 2024).
Read together, the two opinions raise questions that go well beyond simple possession. How much further does Indiana’s statute reach than the federal one? Does the same constitutional shield protect producing/creating obscene virtual images, not simply possessing them? And what happens when the “virtual” image is not purely synthetic — when it starts from an image of a real, identifiable child but is “morphed” into CSAM?
The Anderegg Framework
Anderegg was charged with producing, distributing, transferring, and possessing AI-general CSAM. The government conceded that none of the images depicted an actual child. The district court dismissed only the possession count, and the Seventh Circuit affirmed, walking through three U.S. Supreme Court cases in sequence:
Stanley v. Georgia (1969): protects the private, in-home possession of obscene material from criminal prosecution; the government’s police power over obscenity “simply does not extend to mere possession by the individual in the privacy of his own home.”
Osborne v. Ohio (1990): carved out an exception for actual CSAM, because the government’s interest there is not paternalistic control over what a person thinks but protection of real child victims by drying up the market that harms them.
Ashcroft v. Free Speech Coalition (2002): virtual CSAM (which is not produced with a real child) does not fall under Osborne’s exception because there is no actual victim and no intrinsic link to child sexual abuse.
Because Anderegg’s images were obscene, home-possessed, and untethered to an actual child, the Seventh Circuit concluded Stanley controlled and Osborne’s exception did not apply. The concurring judge all but invited the U.S. Supreme Court to revisit the issue. But for now, the rule stands.
Indiana Got There First…Sort Of
Indiana practitioners may recognize this fact pattern. In F.G. v. State, the Court of Appeals reversed the denial of a motion to dismiss two Level 6 felony possession counts involving manga-style “hentai” images. The images were not AI-generated but hand-drawn. The State conceded that the images did not involve real children. Applying the same framework applied in Anderegg, the Court of Appeals reversed.
Does the Same Logic Protect Production?
Both Anderegg and F.G. were scrupulously narrow, addressing only possession of virtual CSAM. But could the same argument be made for creating/producing virtual CSAM, which still does not involve real child victims? Stanley’s entire foundation is locational: the right to possess obscene material privately, in your own home, is protected because the Court held that allowing the government to cross the threshold of one’s home would be akin to policing people’s thoughts. When Stanley was decided, production would have required real child victims and items that moved through various channels of interstate commerce (cameras, lighting, etc.). But creating of virtual CSAM requires only a computer or, under Indiana law, simply a pen and paper.
The Morphing Problem: When Virtual is Not Totally Virtual
The discussion thus far assumes the image is genuinely synthetic; no real child’s likeness was involved in creating it. But “morphed” CSAM is different, as a real child’s image is utilized. The U.S. Supreme Court’s most recent opinion, Free Speech Coalition, mentioned morphed CSAM but expressly declined to address the issue.
Indiana practitioners may soon have to, though, as the issue is already showing up in charging documents. Last month Indianapolis police arrested a photographer who had used innocuous images of real children to create morphed CSAM through generative AI.
The technology is moving faster than legal doctrine. This is not a novel problem. For criminal practitioners, that is not a cause for concern. It is a strong reason to stay informed on evolving technology and to think creatively about how current (perhaps outdated) legal doctrine applies to this new technology.