Alex Bigman
In December, New York amended its postmortem publicity rights statute to expand civil liability for the unauthorized use of “digital replicas.”1 The Empire state thereby joined a parade of others—including Illinois, Tennessee, and California—that have been busy fortifying publicity rights for the age of A.I.2 Historically, New York has taken a more cautious approach to publicity rights than California, and the new amendments prove no exception. This time, though, the more striking distinction lies not between East Coast and West, but rather between New York’s relatively moderate statute and the far more sweeping legislation—including the NO FAKES Act and the NO AI FRAUD Act—currently pending before the U.S. Congress.
As with A.I. itself, the impact of this policymaking flurry is hard to predict. Publicity rights proponents will find much to celebrate in it. Free speech advocates will be putting the proposed and enacted language under a microscope. Close inspection reveals subtle but telling distinctions that could portend vastly different futures for the landscape of media law.
To start, take the New York statute. The state’s pre-existing provisions for postmortem publicity rights—which themselves date back only to 2020—afforded deceased performers and celebrities protection only against the commercial exploitation of their names, likenesses, and voices.3 The new amendment does away with the commerciality requirement where digital replicas are concerned: knowingly using one of these A.I. products is now unlawful in any “audiovisual work, sound recording, or … live performance of a musical work.”4 With some important exceptions.
A blanket proscription of A.I. simulation, no matter the use or context, would likely fall afoul of the First Amendment. New York’s statute accordingly includes a fairly robust free speech carveout covering (1) “parody, satire, commentary, or criticism;” (2) “works of political or newsworthy value, or similar works, such as documentaries, docudramas, or historical or biographical works, including some degree of fictionalization;” (3) “a representation of a deceased performer as themself, including some degree of fictionalization, except in a live performance of a musical work;” and (4) “de minimis or incidental uses.”5
California’s carveout is a bit charier of ceding celebrity image control. Unlike New York’s law, it declines to extend protection to any “representation of the deceased performer as themself.” It also withholds protection from documentary or biographical uses that are “intended to create, and do create, the false impression that the work is an authentic recording in which the individual participated.”6
Now compare these regimes to the proposed federal statutes, both of which purport to grant all Americans—even the non-celebrities among us—a “property right” in their names, likenesses, and voices.7 The NO FAKES Act creates exceptions resembling California’s, but further narrows its free speech protections by requiring that a digital replica used in news programming be “the subject of, or … materially relevant to, the subject of that broadcast or account,” and that works of commentary, criticism, satire, or parody be “consistent with the public interest” in “bona fide” uses of those genres.”8
The NO AI FRAUD Act takes a different tack. It provides that “First Amendment protections shall constitute a defense to an alleged violation,” then states that “in evaluating any such defense, the public interest in access to the use shall be balanced against the intellectual property interest in the voice or likeness,” considering factors such as “whether (1) the use is commercial, (2) the individual whose voice or likeness is at issue is necessary for and relevant to the primary expressive purpose of the work in which the use appears, and (3) the use competes with or otherwise adversely affects the value of the work of the owner or licensee of the voice or likeness rights at issue.”9
This balancing test resembles the Copyright Act’s fair use provision, but with important differences. First, it omits the Copyright test’s third factor—the “amount and substantiality of the portion used”—thus depriving defendants of the de minimus defense that New York, California, and the NO FAKES Act all preserve.10 Second, it requires that any digital replica be “necessary for and relevant to the primary expressive purpose of the work,” a burdensome element that would require judges to assume the role of art critics.
Such fair use-like exemptions are likely sufficient to render the new and proposed laws facially constitutional, but ultimately, these provisions are only guidelines for the judiciary: it will fall to the courts to decide whether a given defendant’s conduct lies within the zone of protection provided by the First Amendment. It is not difficult to imagine A.I.-assisted replicas with a credible claim to being parody, commentary, fictionalized docudrama, de minimus uses, or simply works of art. How will courts apply existing publicity rights law—a notoriously messy area—to such technologically novel situations?11
[1] N.Y. Civ. Rights Law § 50-f (McKinney 2024).
[2] See, e.g., Tenn. Pub. Acts ch. 588 (2024) (amending Tenn. Code Ann. §§ 47-25-1101 et seq. (2024)) (Ensuring Likeness, Voice, and Image Security Act of 2024, effective July 1, 2024).
[3] See N.Y. Civ. Rights Law § 50-f(2)(a).
[4]N.Y. Civ. Rights Law § 50-f(2)(b).
[5] N.Y. Civ. Rights Law § 50-f(2)(d)(ii).
[6] Cal. Civ. Code § 3344.1.
[7] H.R. 2794, 119th Cong. (2025–2026) (NO FAKES Act of 2025); H.R. 6943, 118th Cong. (2023–2024) (No Artificial Intelligence Fake Replicas and Unauthorized Duplications Act of 2024) (“No AI FRAUD Act”).
[8] H.R. 2794, 119th Cong. (2025–2026), 19–20.
[9] H.R. 6943, 118th Cong. (2023–2024), 10.
[10] For the Copyright Act’s fair use test, see 17 U.S.C. § 107.
[11] On the messiness of First Amendment jurisprudence as applied to publicity rights law, see Robert C. Post & Jennifer E. Rothman, The First Amendment and the Right(s) of Publicity, 128 Yale L.J. 86, 125 (2018).