Home Possession of AI-Generated Porn Depicting Fake Children Remains Constitutionally Protected, Seventh Circuit Rules

From U.S. v. Anderegg, decided Tuesday by Seventh Circuit Judge John Lee, joined by Judges Doris Pryor and Joshua Kolar:

Steven Anderegg is charged with producing, distributing, and possessing visual depictions of minors engaged in sexually explicit conduct and transferring such material to a minor under the age of sixteen. According to the government, Anderegg produced these images using Stable Diffusion, a generative artificial intelligence (“GenAI”) software that allowed him to create hyper-realistic images of prepubescent children engaging in sexually explicit acts.

The appeal in this case deals only with Anderegg’s challenge to the home possession charge, not to the production or distribution charge. Here’s the court’s analysis:

Obscenity, along with certain other categories of speech, has historically fallen “outside the scope of the First Amendment.” And, although the precise meaning of “obscenity” is difficult to pin down, the Supreme Court has offered the following articulation: “[S]peech is obscene to the public at large—and thus proscribable—if (a) ‘the average person, applying contemporary community standards[,] would find that the work, taken as a whole, appeals to the prurient interest’; (b) ‘the work depicts or describes, in a patently offensive way, sexual conduct specifically defined by the applicable state law’; and (c) ‘the work, taken as a whole, lacks serious literary, artistic, political, or scientific value.'” Miller v. California (1973).

At the same time, “[t]he Constitution extends special safeguards to the privacy of the home.” The Supreme Court addressed the clash of these two principles—the government’s interest in proscribing obscenity and an individual’s privacy in one’s home—in Stanley v. Georgia (1969)…. Acknowledging that the government retains “broad power to regulate obscenity,” the Supreme Court held that the State’s “power simply does not extend to mere possession by the individual in the privacy of his own home.” “If the First Amendment means anything,” the Court explained, “it means that a State has no business telling a man, sitting alone in his own house, what books he may read or what films he may watch.” …

Approximately a decade after Stanley, the Supreme Court addressed whether the promotion and sale of CSAM [child sexual abuse material] implicated First Amendment rights. [In N.Y. v. Ferber 1982),] the Court nevertheless held that “[t]he test for child pornography is separate from the obscenity standard enunciated in Miller” because CSAM was different from adult obscene materials in several important ways. First, the state’s interest in protecting the physical and psychological well-being of minors and preventing the sexual exploitation and abuse of children “constitute[d] a government objective of surpassing importance.” Second, the distribution of CSAM was “intrinsically related to the sexual abuse of children” because the materials “are a permanent record of the children’s participation and the harm to the child is exacerbated by their circulation.” Third, the advertising and selling of CSAM “provide[d] an economic motive for and are thus an integral part of the production of such materials.” Fourth, the value of performances and photographs of children engaged in lewd sexual conduct was “exceedingly modest, if not de minimis.” … Eight years later, the Supreme Court [held that] the protections articulated in Stanley would [not] extend to the possession of CSAM in one’s home. Osborne v. Ohio (1990).

[Finally, i]n Ashcroft v. Free Speech Coalition (2002), the Supreme Court confronted the question of whether virtual CSAM, like actual CSAM, was beyond the First Amendment’s reach…. “By prohibiting child pornography that does not depict an actual child,” the Court stated, “the statute goes beyond [Ferber] which distinguished child pornography from other sexually explicit speech because of the State’s interest in protecting the children exploited by the production process.” … For much the same reason, the Court remarked, Osborne too was distinguishable, because it had “anchored its holding in the concern for the participants, those whom it called the ‘victims of child pornography.'” In Osborne, the state had a compelling interest in protecting children by attempting to eliminate CSAM at all levels of the distribution chain and preventing CSAM from being used to solicit minors. This was not the case in Free Speech Coalition, the Supreme Court declared, because the depictions did not involve actual children….

[T]he government [argued, among other things,] that, because it was often hard to distinguish virtual CSAM from real CSAM, prosecuting those who produced such material using actual children was very difficult. But the Court made short shrift of this argument, declaring that “[t]he Government may not suppress lawful speech as the means to suppress unlawful speech. Protected speech does not become unprotected merely because it resembles the latter.”

After Free Speech Coalition, Congress went back to the drawing board and enacted the statute at issue here. The Prosecutorial Remedies and Other Tools to end the Exploitation of Children Today Act of 2003 (“PROTECT Act”) included a new provision which, in part, proscribed possession of “a visual depiction of any kind … that (1)(A) depicts a minor engaging in sexually explicit conduct; and (B) is obscene.” … [I]t expressly does not require that “the minor depicted actually exist.” …

Because the images for which Anderegg is charged do not depict an actual child, Ferber and Osborne do not directly apply to this case…. And, because the arguments the government advances to remove the materials in question from Stanley‘s purview have all been squarely rejected by the Supreme Court in Free Speech Coalition, we conclude that § 1466A(b)(1) is unconstitutional as applied to Anderegg…..

Judge Lee, joined by Judge Kolar, also wrote a concurring opinion “to note that, given 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”:

[N]early twenty five years after the Court’s decision, Free Speech Coalition controls in a day where “AI-generated CSAM is virtually indistinguishable from CSAM created through the physical abuse of a child.” At the time of Free Speech Coalition, several Justices predicted as much. Justice O’Connor echoed many of the concerns the government argues today, namely that “rapidly advancing technology soon will make it all but impossible to” enforce prohibitions of actual CSAM and “[c]omputer-generated images … bear a remarkable likeness to actual human beings.”

Justice Thomas too forecasted that “technology may evolve to the point where it becomes impossible to enforce actual child pornography laws because the Government cannot prove that certain pornographic images are of real children.” In his view, “if technological advances thwart prosecution of ‘unlawful speech,’ the Government may well have a compelling interest in barring or otherwise regulating some narrow category of ‘lawful speech’ in order to enforce effectively laws against pornography made through the abuse of real children.”

According to some experts, that day may have arrived. Recent data suggests that the influx of AI-generated images reported to the NCMEC CyberTipline imposes increasing burdens on law enforcement…. “Even the images that do not depict a real child put a strain on law enforcement resources and impede identification of real child victims.” … [P]erpetrators may “disguise pictures of real children being abused by making the image look computer-generated” ….

What is more, AI-generated CSAM may also present other challenges not raised by the government in this appeal. For example, researchers have found that “a major dataset used for training image-generating AI contained hundreds of CSAM images.” Thus, unlike the virtual CSAM at issue in Free Speech Coalition, the market for AI-generated virtual CSAM images may impact the demand for actual CSAM on which the AI models can train. Such a finding would implicate many of the concerns raised in Osborne.

In this case, the record is insufficient to evaluate these complex issues, and, of course, we are bound by the dictates of Free Speech Coalition. But it goes without saying that “[t]he prevention of sexual exploitation and abuse of children constitutes a government objective of surpassing importance.” The stakes are high, and we must be mindful of the ever-accelerating advancements in image generation technology.

Joseph Aragorn Bugni represents Anderegg.