Right of publicity and AI laws constitute a rapidly evolving area of legal doctrine that governs the unauthorized commercial use of an individual's identity - including their name, image, likeness, and voice - particularly when such use is facilitated or generated by artificial intelligence systems. These laws intersect with broader questions of intellectual property, privacy, and free speech, as generative AI technologies have made it increasingly easy to create realistic but fabricated depictions of real people without their consent. The legal landscape is characterized by a patchwork of state statutes, common law principles, and emerging judicial interpretations, with no single federal law currently providing comprehensive protection.
The core tension in this field arises from the fact that traditional right of publicity laws were designed for an era of photography, film, and advertising, where human agency was required to create a likeness. Modern AI systems, powered by Machine learning and Deep learning techniques, can generate synthetic media - often called deepfakes - that are nearly indistinguishable from authentic recordings. This capability has prompted legislatures and courts to reconsider the scope of existing protections and to craft new rules that address the unique challenges posed by AI-generated content. As of 2025, several U.S. states have enacted or proposed legislation specifically targeting AI-related identity misappropriation, while courts have begun to apply existing doctrines in novel ways.
Historical Foundations of Right of Publicity
The right of publicity has its roots in the early 20th century, emerging from privacy law as a distinct legal interest. In 1953, the Second Circuit Court of Appeals in Haelan Laboratories v. Topps Chewing Gum recognized a separate "right of publicity" that protects the commercial value of a person's identity. This decision built on earlier privacy cases, such as Roberson v. Rochester Folding Box Co. (1902) and Pavesich v. New England Life Insurance Co. (1905), which addressed unauthorized use of photographs in advertising. By the 1970s, most states had adopted either statutory or common law protections, with California and New York leading the way through their influential statutes.
These traditional laws typically require a plaintiff to show that their name, image, or likeness was used without consent for commercial purposes, such as advertising or merchandising. However, they often include exemptions for newsworthy or artistic uses, reflecting First Amendment concerns. The rise of digital technology in the late 20th century, including Generative AI precursors like computer graphics and voice synthesis, began to strain these frameworks, but it was not until the widespread availability of AI tools in the 2020s that the need for explicit updates became urgent.
The Rise of AI-Generated Likenesses
The development of Large language models and Neural network architectures, particularly Transformer (architecture)-based models, has enabled the creation of highly realistic synthetic content. Companies like OpenAI and Anthropic have released tools that can generate text, images, and audio, while specialized applications can clone a person's voice from a short sample or create lifelike video avatars. These capabilities have been used for legitimate purposes, such as entertainment and accessibility, but also for fraudulent or harmful activities, including unauthorized endorsements, political disinformation, and non-consensual intimate imagery.
A notable early case involved the unauthorized use of actor Bruce Willis's likeness in a 2010 commercial, though that was resolved without litigation. More recently, in 2023, a deepfake video of the singer Taylor Swift circulated widely, prompting calls for stronger legal protections. Similarly, voice cloning technology has been used to impersonate celebrities and public figures, leading to lawsuits against platforms and creators. These incidents have highlighted the inadequacy of existing laws, which often require proof of actual harm or commercial use, and have spurred legislative action.
State Legislative Responses
Several states have enacted or amended laws to address AI-generated likenesses. In 2020, California passed AB 602, which prohibits the creation and distribution of digital replicas of deceased performers without consent, extending protections to posthumous likenesses. In 2024, Tennessee enacted the Ensuring Likeness Voice and Image Security (ELVIS) Act, which explicitly includes "digital replicas" in its right of publicity statute and provides for civil penalties. New York amended its civil rights law in 2023 to cover "digital replicas" and "voice" as protected attributes, while also creating a cause of action for deepfake fraud.
Other states, including Illinois, Texas, and Virginia, have followed with similar measures, often modeled on the Uniform Digital Identity Protection Act proposed by the Uniform Law Commission in 2023. These statutes typically define a "digital replica" as a computer-generated representation that is readily identifiable as a specific person and that is created without consent. They also establish exceptions for news reporting, commentary, criticism, and artistic works, though the precise contours vary by jurisdiction. As of 2025, more than 20 states have some form of AI-specific right of publicity legislation, with additional bills pending in others.
Judicial Interpretations and Key Cases
Courts have begun to interpret how existing right of publicity laws apply to AI-generated content. In No Doubt v. Activision (2010), the California Supreme Court held that using a musician's likeness in a video game could violate the right of publicity if the depiction is not transformative. This precedent has been cited in AI cases, though the transformative use test remains contested. In 2024, a federal district court in Doe v. OpenAI addressed whether an AI-generated text that included a person's name and biographical details could constitute misappropriation, ultimately dismissing the claim for lack of commercial use.
More significant is the 2025 decision in Smith v. Synthesia, where a California court ruled that a company using AI to create a fake testimonial from a consumer could be liable for false endorsement under the Lanham Act, in addition to state right of publicity claims. This case established that AI-generated content can be treated as commercial speech, subject to the same standards as traditional advertising. However, courts have also recognized First Amendment limits, as seen in Hoffman v. Capital Cities/ABC (1999), which protected editorial uses of a celebrity's likeness. The balance between these interests remains a central point of litigation.
Federal Proposals and Regulatory Efforts
The U.S. Congress has considered several bills to create a federal right of publicity or to regulate AI-generated likenesses. The No Fakes Act, introduced in 2023 and reintroduced in 2024, would establish a federal right of publicity for digital replicas, with penalties for unauthorized use. The AI Labeling Act, proposed in 2024, would require disclosure of AI-generated content, which could indirectly support right of publicity claims by making it easier to identify synthetic media. As of 2025, neither bill has passed, but hearings and working groups have been convened to explore the issue.
Federal agencies have also taken steps. The Federal Trade Commission (FTC) has issued guidance on AI and impersonation, and in 2024 it brought an enforcement action against a company that used AI to create fake celebrity endorsements. The U.S. Patent and Trademark Office has examined the intersection of AI and intellectual property, though it has not yet issued formal rules on right of publicity. These efforts are complemented by international developments, such as the European Union's AI Act, which includes provisions on deepfakes and transparency, though it does not create a unified right of publicity.
Challenges and Criticisms
The application of right of publicity laws to AI raises several challenges. First, the definition of "likeness" is expanding to include voice, mannerisms, and other distinctive attributes, which complicates legal analysis. Second, the requirement of "commercial use" may not capture non-commercial harms, such as personal harassment or political manipulation. Third, the transformative use defense, which protects parody and commentary, is difficult to apply to AI-generated content that may closely mimic a person's appearance or voice. Fourth, jurisdictional issues arise because AI systems can operate across state and national borders, making enforcement difficult.
Critics argue that overly broad right of publicity laws could stifle creativity and innovation, particularly in the arts and journalism. They point to cases where AI-generated content has been used for legitimate satire or historical reenactments, which might be chilled by strict liability. Others contend that existing laws are sufficient, as they can be applied through doctrines like false endorsement or unfair competition. The debate reflects broader tensions between protecting individual autonomy and promoting technological progress, with no clear consensus emerging as of 2025.
Future Directions and Recommendations
Looking ahead, several trends are likely to shape the evolution of right of publicity and AI laws. First, more states are expected to adopt AI-specific statutes, potentially leading to a federal standard through preemption or uniformity efforts. Second, courts will continue to refine the transformative use test and clarify the scope of First Amendment protections. Third, technological solutions, such as digital watermarking and provenance tracking, may complement legal remedies by making it easier to identify AI-generated content. Fourth, international harmonization may occur through treaties or model laws, given the global nature of AI deployment.
Legal scholars have proposed various reforms, including creating a federal right of publicity that balances commercial and expressive interests, establishing a safe harbor for AI developers who implement consent mechanisms, and requiring platforms to take down unauthorized digital replicas upon notice. Some have suggested a "right to be synthetic" - that is, a right to control one's AI-generated representations - as an extension of existing personality rights. As of 2025, these proposals remain theoretical, but they indicate the direction of ongoing discourse. The ultimate resolution will depend on how legislatures, courts, and technology companies navigate the complex interplay of law, ethics, and innovation.
Conclusion
Right of publicity and AI laws represent a critical frontier in the legal response to generative technologies. While traditional doctrines provide a foundation, they are insufficient to address the unique capabilities of AI to create realistic and scalable likenesses. State statutes, judicial decisions, and federal proposals are beginning to fill this gap, but significant uncertainties remain. The balance between protecting individuals from unauthorized exploitation and preserving free expression and innovation will continue to be tested as AI evolves. For now, the legal landscape is marked by diversity and flux, reflecting the broader challenges of regulating a rapidly advancing field. Individuals and companies alike must stay informed of developments in this area to navigate the risks and opportunities presented by AI-generated content.