By: Dillon Koch

If you spend a few minutes scrolling through TikTok, you’re sure to come across Plankton belting out an 80s pop hit, or Harry Styles duetting a Taylor Swift song (ostensibly written about their iconic fling). However, the Bikini Bottom villain and Dunkirk’s bravest soldier never actually made those videos – they were created using generative Artificial Intelligence (AI).

With the rise in accessible AI tools, effectively anyone has the power to make anyone say anything—just ask Tom Hanks and Gayle King, who were victims of generative AI when they “appeared” in commercials for dental services and weight loss products. However, neither Hanks nor King ever filmed, recorded, or otherwise approved either campaign. While generative AI can lead to harmless viral content on social media, it can also cause real issues for performers who never authorized these performances.  

Producers Proposed Unlimited Access to Background Actors

In what seemed to be the nexus for the Screen Actors Guild – American Film, Television, and Radio Actors (SAG-AFTRA) strike, initial bargaining proposals from the Alliance of Motion Picture and Television Producers (AMPTP) indicated that studios were interested in minimizing costs by maximizing how background actors would be cast, compensated… and created.

According to union members, AMPTP proposed terms that would allow them to have nearly unlimited use of low-cost background actors. AMPTP proposed that SAG-AFTRA background performers should be able to be scanned, get a single day’s pay, and their companies could own that scan, the performer’s image, and their likeness. With this, AMPTP sought to be able to use the performer’s scan and likeness for an unlimited period, without explicit consent or compensation for individual project usage. This could have a significant impact on financial stability and workforce participation for union members, and has led to the longest strike in SAG-AFTRA’s history.

Now, in an effort to preserve their craft while also maintaining the ability to work and survive, members of SAG-AFTRA aim to obtain similar AI protections to the Writers’ Guild of America (WGA): putting creative humans at the forefront of film and television, and leveraging AI as a tool rather than an inherent creative force.

Popular Copyright Theory Is Favorable Toward Performers, But Jurisprudence Disagrees

Prominent theories of copyright seem to indicate that performers should be entitled to rights over their likeness and expression of themselves. Hegelian and Kantian perspectives indicate that personhood, autonomy, and self-expression should be protected by copyright. The intuition is that the best way of providing control over resources is to recognize property

rights, and a person’s property interest is strongest in the resources that entangle their personality. An extension of this theory posits that property provides a mechanism for self-definition, personal expression, and dignity of an individual person.

Even the definition of copyright itself “includes” well-established categories of eligible works, indicating that there is room to expand copyrightability beyond the named eight categories as interpretation allows. While jurisprudence has not revealed new categories in the definition of “works,” the drafters of the Constitution seemed to acknowledge that creativity’s broad scope could lead to unforeseen innovative creations.

While it seems like a no-brainer that a performer should maintain rights to their likeness and expression, courts have established clear jurisprudence to the contrary. In the seminal case Garcia v. Google, courts held that a performance was not copyrightable by the performer, which which resulted in a plaintiff’s likeness being used in anti-Muslim propaganda without her explicit knowledge or consent. The dissent believed that Garcia was effectively “bamboozled, acknowledging that its decision in this case “robs performers and other creative talent of rights Congress gave them.” However, the Ninth Circuit ultimately held that, although the film’s treatment of Garcia was “blasphemous,” her performance was not substantially creative enough to constitute authorship. Even if she were granted authorship, she was not responsible for the fixation of her performance, and therefore could not get copyright protection.

So, despite the philosophy that performers should be able obtain copyrights in order to protect their personhood and rights to self-expression, courts have maintained that performers are generally not eligible for copyrights to their performance. However, with increasing weariness around generative AI, this may be the perfect moment for Congress and the courts to add an additional category to copyrightable works: likeness.

Introducing a New Category of Copyrightable Material

In a seemingly rare act of bipartisanship, four members of Congress have drafted the Nurture Originals, Foster Art, and Keep Entertainment Safe (NO FAKES) Act. Senators Coons, Blackburn, Klobuchar, and Tillis aim to combat AI’s “unique challenges that make it easier than ever to use someone’s voice, image, or likeness without their consent” through policies regulating the use and impact of generative AI. The Senators aim to defend individual rights and abide by First Amendment rights to free speech, without stifling innovation and creativity.

The NO FAKES Act would prevent a person from producing or distributing an unauthorized AI-generated replica of an individual to perform in a recording without the explicit consent of the individual being replicated. The person creating or sharing the replication, as well as the platform knowingly hosting it, would be liable for the damages caused by the AI-generated fake. 

SAG-AFTRA has applauded the proposal, acknowledging that the rise in accessible generative AI tools has increased opportunity to exploit the voices and likenesses of union actors without their knowledge, consent, or fair compensation. Especially after WGA reached their resolution with studios, SAG-AFTRA seems more hopeful than they were at the beginning of the summer. If WGA can receive copyright and employment protections from AI-generated content, why shouldn’t performers be entitled to the same?

It seems, with the proposal of NO FAKES, Congress hopes to increase protections around generative AI by broadening copyright-like protections to to include performers’ likenesses. As SAG-AFTRA President Fran Drescher noted, “A performer’s voice and their appearance are all part of their unique essence, and it’s not ok when those are used without their permission,” which is directly aligned with copyright principles of protecting personhood and self-expression. 

While the bill doesn’t mention copyright specifically, NO FAKES sends a signal that likeness and performance are essential creative works which should receive federal protection. This seems likely to have a direct impact on future copyright infringement claims. As more actors file claims against unauthorized use of their likeness, government leaders are eager to impose uniform limitations on AI. The NO FAKES bill ultimately provides the perfect opportunity to protect autonomy and creativity, and in such finally aligning with principles of copyright by expanding creative protection to actors.