Abstract
Advances in artificial intelligence (AI) have enabled the creation of hyper-realistic child sexual abuse images without any real child involved. These AI-generated depictions pose a unique challenge: they replicate the most pernicious form of pornography—child pornography—while potentially being claimed as “speech” immune from regulation. This paper argues that it is both constitutionally permissible and morally imperative to criminalize AI-generated child pornography. This article presents two doctrinal pathways to justify a ban. First, such content can be treated as unprotected speech under existing First Amendment exceptions. Courts have long held that child pornography and obscenity lie outside First Amendment protection, and a carefully defined prohibition on AI-synthesized child sexual abuse material can fit within these categories or be analogized to contraband, thus sidestepping strict scrutiny. Second, even if considered protected expression, a narrowly tailored ban can survive strict scrutiny by advancing the compelling interest of protecting children and society from sexual exploitation. This paper demonstrates that lawmakers can define the offense to cover only realistic, valueless depictions virtually indistinguishable from real child pornography, thereby targeting the core harm while avoiding overbreadth. This article concludes that a well-crafted prohibition on AIgenerated child porn not only withstands constitutional scrutiny but is a necessary extension of child protection into the digital age.
Recommended Citation
Adam Abdin, Virtual Harm, Real Consequences: Constitutional Justifications for Criminalizing AI-Generated Child Pornography, 20 FIU L. Rev. 1345 (2026), https://doi.org/10.25148/lawrev.20.4.12.
Included in
Constitutional Law Commons, Criminal Law Commons, First Amendment Commons, Privacy Law Commons, Sexuality and the Law Commons
