Voice, Likeness, and Fair Use in the Age of AI 

As artificial intelligence (AI) continues to develop, it is increasingly intersecting with human identity, particularly the replication of voice and likeness. These advancements have raised complex legal questions about ownership, consent, and protection in the digital age. Recent public controversies involving AI-generated voices resembling high-profile individuals, including Taylor Swift, Tom Hanks, and Scarlett Johansson, have brought national attention to these unresolved issues. 

In May 2024, Scarlett Johansson publicly objected to OpenAI’s use of a voice for its ChatGPT assistant “Sky,” which she alleged bore a strong resemblance to her own. The incident drew parallels to the 2013 film Her, in which Johansson voiced a fictional AI. A social media post by OpenAI’s CEO referencing the word “her” when “Sky” was released further fueled public scrutiny. OpenAI subsequently paused the use of “Sky” and clarified that the voice actor was not intended to imitate Johansson. 

This incident illustrates the legal uncertainty surrounding AI-generated content that mimics real people (intended or otherwise). In particular, it raises an important question: When does the use of AI cross the line from lawful expression into unlawful appropriation? 

Existing laws already apply to AI, including intellectual property theories of copyright and right of publicity.

Copyright Law 

Copyright is a form of intellectual property that protects original works of authorship once they are fixed in a tangible medium of expression. Current U.S. copyright law offers limited protection for voice and likeness. 

Earlier this year, in Thaler v. Perlmutter, the D.C. Circuit Court of Appeals reaffirmed that copyright protection requires human authorship. This means that works created entirely by AI, without meaningful human involvement, are not eligible for protection. This aligns with the recent U.S. Copyright Office guidance, which emphasizes that authorship must reflect human creativity to qualify for copyright protection. 

Voice and likenesses are generally not protectable under copyright law unless they are part of a larger copyrighted work, such as a film or song. (Even in this circumstance, the voice/likeness isn’t protected, just the particular instance of it in the film/song). This limits the ability of individuals to assert copyright claims against unauthorized AI-generated imitations of their voices or appearances. As explained below, there might be other potential claims. 

The fair use doctrine allows limited use of copyrighted materials without permission for purposes such as commentary, criticism, and parody. Courts assess fair use based on four factors, codified at 17 U.S.C. § 107: 

1. The purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; 
2. The nature of the copyrighted work; 
3. The amount and substantiality of the portion used in relation to the copyrighted work as a whole; and 
4. The effect of the use on the potential market for or value of the copyrighted work. 

Some AI-generated works that imitate real individuals may qualify as fair use. However, courts have yet to define how fair use applies to synthetic media, particularly when it involves non-copyrightable elements like voice and likeness. This uncertainty highlights a growing legal tension between safeguarding individual identity and protecting expressive freedoms under the First Amendment. 

Right of Publicity

The right of publicity protects individuals against unauthorized commercial use of their identity, including their name, likeness, and, in some cases, voice. However, the right of publicity is governed by various state laws, in those states that have a right of publicity, with no comprehensive federal statute currently in place. 

To bring a successful right of publicity claim, a plaintiff typically must show that (1) their identity is identifiable, (2) the use was unauthorized, (3) the use was commercial, and (4) they suffered monetary harm. Precedents such as Midler v. Ford Motor Co. and Waits v. Frito-Lay demonstrate that distinctive imitations of a person’s voice can be actionable, particularly when used in commercial advertising. However, these cases primarily protect well-known figures, leaving most individuals with fewer options for recourse. 

Recent Legislative Developments

While federal legislation has not advanced, states have increasingly taken the lead in addressing the challenges posed by AI-generated impersonations. All 50 states recognize common law rights against the unauthorized use of a person’s name, image, or likeness, and around 30 states provide statutory rights of publicity and privacy. States including California, Tennessee, and Washington directly regulate digital replicas under these laws. And in 2025 alone, 13 additional states have introduced legislation targeting digital replicas.

California, for example, has passed several relevant laws, including AB 1836, which extends post-mortem publicity rights to cover AI-generated replicas, and AB 2602, which requires informed consent for the use of digital replicas in entertainment contracts. Arkansas has similarly amended the Frank Broyles Publicity Rights Protection Act to address AI-generated voice and likeness protections. As policymakers consider new regulations to address synthetic media and AI-driven impersonation, it’s important to recognize that many legal tools already exist and are evolving. However, the increased number of state laws may still create challenges for developers, platforms, and rights holders navigating inconsistent standards across jurisdictions. 

Conclusion

As AI capabilities continue to evolve, the legal landscape surrounding voice and likeness rights remains unsettled. Existing doctrines — including copyright, fair use, and the right of publicity — provide some tools for addressing these issues, and can adapt to new technologies, even though they may not have been designed with synthetic media in mind. 

Going forward, policymakers and courts face the challenge of balancing individual rights against the principles of free expression and innovation. A clear, consistent legal framework — one that targets liability toward bad actors who intentionally or knowingly engage in deceptive acts using a digital replica, rather than the tools or services that merely enable their creation or distribution — will be essential. Such an approach can help protect personal identity, free expression, and the responsible development of AI technologies.