A Real Account of Deep Fakes

Abstract

Laws regulating pornographic deepfakes are usually characterized as protecting privacy or preventing defamation. But privacy and defamation laws paradigmatically regulate true or false assertions of fact about persons. Anti-deepfakes laws do not: the typical law bans even media that no reasonable observer would understand as factual. Instead of regulating statements of fact, anti-deepfakes laws ban outrageous depictions as such. This is an unrecognized departure from established privacy and defamation law, and it carries serious constitutional stakes. Because anti-deepfakes laws ban outrageous imagery irrespective of any factual assertions it makes, the rationales for defamation and almost all privacy doctrines do not justify these statutes under the First Amendment. Properly understood, anti-deepfakes laws fall into a distinct, and constitutionally disfavored, category: laws that forbid expression because it is offensive.

This Article conducts the first scholarly survey of every anti-deepfakes statute and distills the typical law. It then uses semiotic theory to explain how deepfakes differ from the media they mimic and why those differences matter legally. Photographs and video recordings record events. Deepfakes merely depict them. Justifications for regulating records do not necessarily justify regulating depictions. Many laws—covering everything from trademark dilution to flag burning to “morphed” child sexual abuse material (CSAM)—have banned offensive depictions as such. Several remain in effect today. Yet when such bans are challenged, courts mischaracterize imagery to sidestep constitutional scrutiny: courts pretend fictional depictions are factual records. Anti-deepfakes laws resist this dodge. The laws will force courts to confront squarely whether a statute may ban offensive expression as such—and thus to determine whether a history and tradition of such regulation can be reconciled with First Amendment jurisprudence that would seem to forbid it.

Keywords: deepfake, privacy, ai, artificial intelligence, pornography, sexual privacy, semiotics, peirce, privacy, speech, First Amendment, expression

JEL Classification: K24, O34, K38, K14, K15