AI-GENERATED DEEPFAKES AND PERSONALITY RIGHTS: A NEW FRONTIER FOR INTELLECTUAL PROPERTY LAW

Аннотация

The rapid evolution of artificial intelligence (AI), particularly in generative media, has given rise to deepfakes — hyper-realistic digital manipulations of a person’s likeness. This phenomenon challenges the boundaries between intellectual property (IP), privacy, and personality rights. While existing copyright law protects expressions of creativity, it does not squarely address the unauthorized use of a real person’s image or voice through AI. This article explores whether and how personality rights can serve as a legal response to the rise of deepfakes, comparing U.S. jurisprudence with emerging perspectives in jurisdictions like Uzbekistan. The study argues for a reconceptualization of legal identity in digital space, proposing strongerintegration between personality rights and IP frameworks. Keywords:Deepfakes, Artificial Intelligence, Personality Rights, Right of Publicity, Digital Identity, Intellectual Property Law, Moral Rights, AI Regulation, Comparative Law, UzbekistanLegal Framework, Synthetic Media, Image Rights, Legal Harm, Privacy 1.Introduction The advent of AI-generated deepfakes has transformed the way we understand identity, creativity, and control over one’s likeness. These synthetic media outputs — often indistinguishable from authentic content — raise critical legal and ethical questions: Who owns the digital representation of a person? Can a manipulated image be copyrighted? Does the subject ofa deepfakepossess legal remedies undercurrent personality rights frameworks? This paper focuses not on the general risks posed by AI, but specifically on the intersection between deepfake technology and personality rights, a field inadequately protected under traditional intellectual property regimes. WhiletheUnited Stateshas developed someprotection under the “right of publicity,” and through tort law such as appropriation and false light, these measures remain fragmented and jurisdiction-specific. In contrast, civil law countries often conceptualize personality rights as moral and inalienable — but how these principles adapt to digital clones remains uncertain.1 The present study seeks to address this gap through a focused, comparative lens — analyzing the legal treatment of AI-generated impersonations and proposing doctrinal innovations to prevent reputational harm, identity theft, and unauthorized commercial exploitation. 1Deepfakes,createdusingGenerativeAdversarialNetworks(GANs),haveblurredthelinebetweenauthenticand syntheticmedia(Chesney&Citron,2019). page100 INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE ISSN:2692-5206,ImpactFactor:12,23 AmericanAcademicpublishers,volume05,issue06,2025 Journal: https://www.academicpublishers.org/journals/index.php/ijai 2.Methods This study adopts a comparative doctrinal methodology, analyzing statutory laws, court decisions, and scholarly commentary from the United States and Uzbekistan, with additional references to European Union developments. Key materials include:  U.S.state-level right of publicitystatutes (e.g., California, New York)  Federal court cases involvingAI impersonation and First Amendment defenses  Uzbek legislation onimage rights and draft discussions ondigital regulation  Legal scholarship onmoral rightsand thedigital self The research is also informed by interdisciplinary sources, including media ethics, digital forensics, and AI policy literature, to account for technological complexity and socio-legal impact. Unlike broader philosophical works, this article emphasizes doctrinal precision and normativereform within existingIP and personality rights frameworks. 3.Results The legal landscape, as currently structured, fails to address the scope and scale of harm caused byAI-generated deepfakes. Key findings from thecomparative analysis include: 3.1U.S.Jurisprudence: Commercial Protection with ConstitutionalLimits In the United States, the right of publicity provides a cause of action against unauthorized commercial use of one’s name, image, or likeness. However, it is not federally codified and varies by state. California's Civil Code §3344 offers a robust framework, but First Amendment challenges — particularly in entertainment and political satire — create legal friction. Courts arereluctant toextend liability to creators ofexpressivecontent, even ifsynthetic. For instance, in Comedy III Productions v. Gary Saderup, the California Supreme Court established the "transformative use" test, where works of art containing celebrity likenesses are protected if they add significant expressive content. This test has been extended to video games (Keller v. Electronic Arts), but how it applies to AI-generated synthetic media remains uncertain. Deepfakes that are hyper-realistic but non-parodic often fall into a gray zone — arguablyexpressive, butreputationally harmful.2 3.2Uzbekistan’s Silence: Absenceof EnforcementTools In Uzbekistan, personality rights exist under general civil law principles, including the Civil Code and the Law on Info