From Short-Form Series to the Entertainment Industry: The Emerging Risks of AI Technology Against Actors’ Publicity Rights

I. Introduction
As new technologies transform the entertainment industry, producers are turning to AI to streamline productions and expand creative possibilities. However, AI’s rapid expansion has also created moral and legal gray areas that policies struggle to keep up with. In July 2026, Business Insider reported that actor Ashley Beloat discovered that a new series recreated her lead in the mini TV show, “Hate the Way I Love You,” using an AI-generated actor in a different series. [1] Although the AI actor looked different, it mimicked Beloat’s movements and voice without her consent. Micro drama producers have even used AI to recreate misleading, sexualized promos to attract new users by using previous actors’ profiles and on-screen work. Transparency and confidentiality have therefore become common challenges in the entertainment field. These cases raise major questions about how to maintain legal and ethical practices that help performers retain control over how producers use their profiles. 

AI companies have started regulating their generative models to mitigate these gray areas. For instance, Zegocloud’s 2023 whitepaper categorized digital human applications into “identity-based” and “service-based” types. [2] Applications with independent identities and personalities are classified as fulfilling “entertainment and social purposes,” while applications considered for their functionality are classified as “replacing real human services.” [3] Despite these regulations, performer-led AI outputs often make it difficult to determine the boundaries between replicating individual identities and producing generic content through Zegocloud’s large language model. 

Although recent legislation and industry agreements have begun to address digital replicas by restricting unauthorized commercial use of individuals’ identities, legislation continues to lag behind rapid technological development, leaving legal frameworks largely reactive rather than proactive in addressing emerging risks in the entertainment field. Thus, existing U.S. publicity, copyright, and contract laws provide important but incomplete protections for actors whose identities are recreated through AI. This article examines how these legal frameworks interact and argues for clearer and more specific consent requirements that protect performers without unnecessarily restricting legitimate creative innovation. 

II. Industry Background: Rise of AI in Short-Form Television 
In today’s fast-paced, short-form video-dominated society, AI short-form series have become the next dominant viewing preference. In the first quarter of 2026, about 128,000 short dramas were released in China, with over 95% containing AI-produced content. [4] Like short-form television producers, producers focusing on other media types in the entertainment field are also looking  into AI tools to lower costs and boost efficiency. Chen Hetian of Vidu AI believes that AI-generated comic content “could generate as much as RMB 50 billion ($7 billion) in revenue this year alone.” [5] Generative models are now capable of generating or altering actors’ faces and voices, digitally modifying existing performances, creating synthetic actors or performers, and reproducing aspects of an existing performance in another production. Producers may choose to adopt this tool to easily localize or dub, or produce additional content from preexisting ones. This is not only problematic in light of the Beloat case; it also affects the entire entertainment industry. For instance, Q’orianka Kilcher sued James Cameron over claims that he used her likeness to create Neytiri in Avatar. [6] Rather than a production tool, AI has now become a substitute for an actor’s identity or performance, creating new challenges to protecting their rights. 

III. The Current Control of Actors’ Identities and Performances from a Legal View

A. Right of Publicity and Digital Replicas 
The right of publicity plays a crucial role in protecting actors’ identities from AI intervention in unauthorized commercial use of their name, image, likeness, voice, or signature. For instance, Tennessee’s 2024 ELVIS Act created “a property right in the use of [an] individual’s name, photograph, voice, or likeness in any medium in any manner.” [7] This means rights holders can sue the direct infringer as well as the company that created the generative AI model. Similar legislation is applied in California’s statutory right of publicity, where using others’ identity “for purposes of advertising or selling, or soliciting purchases of, products, merchandise, goods, or services, without that person’s prior consent” would be considered a violation. [8] These protections are particularly relevant to performers whose recognizable identities can be exploited commercially without their direct participation, safeguarding their rights under close monitoring. 

California has become one of the fastest states in establishing legislation in response to the growing use of AI in the entertainment industry, specifically focusing on digital replicas through technology. The California legislature’s Assembly Bill 1836 defined “digital replica” as “a computer-generated, highly realistic electronic representation that is readily identifiable as the voice or visual likeness of an individual” that is embodied in any form of media and work. [9] Hence, A.B. 1836 mainly provided a private right of action against any person who makes or uses the digital replica of a deceased person’s voice or likeness in other works without any prior consent. [10] This is specifically crucial in monitoring AI generative models, as it focuses not only on living actors who can voice their own rights, but extends the rights to deceased celebrities. 

Similarly, New York civil rights law also has strict restrictions on the right of publicity, although it does not directly include AI-specific regulations. New York Civil Rights Law §§ 50 and 51 state that any person whose identity is used for “advertising purposes or for the purposes of trade without the written consent first obtained” can bring an action in the Supreme Court seeking injunctive relief, compensatory damages, or exemplary damages. [11] Although these provisions were not originally designed specifically for generative AI, they are relevant to assessing certain commercial uses of AI-generated replicas. Across these jurisdictions, publicity-rights laws and related digital-replica legislation provide performers legal avenues to challenge specified unauthorized uses of their identities. 

B. Contract Consent and the Scope of Production Agreement 
Contractual consent and production agreements have similarly become a heavily emphasized area in newly established state legislation regarding AI generative models. Alongside A.B. 1836, California’s A.B. 2602 imposes restrictions on service contracts that allow the creation and use of digital replicas to replace works that an individual should have completed themselves. [12] Within this, a digital replica is only seen as reasonable when individuals sign contracts with informed consent on the intended use of a digital replica with a lawyer or labor union representative. [13] This protects actors from a more practical view and prevents immoral contractual situations where actors become bound by contractual agreements despite violations of their own rights. In New York, Labor Law § 1035(10) states that model management companies are not allowed to create, alter, or manipulate the model’s digital replica through generative AI models without gaining clear written informed consent. [14] As one of the first New York statutes to regulate AI digital replicas, it shows how legislation has heightened consent standards in governing actors’ rights. [15] These state laws ensure that performers retain greater control over how their identities are reproduced by generative AI in contracts, expanding their digital replica rights from a practical industry perspective.

C. Copyright in the Original Production 
Laws and regulations regarding copyright issues have not been expanded to AI-specific restrictions but still remain relevant and effective in the entertainment industry. This can be seen in the Copyright Act of 1976, also known as Title 17 of the United States Code. § 102 states that original audiovisual works are protected, but not abstract ideas, while § 106 grants the copyright owner the exclusive and independent legal right to reproduce, adapt, distribute, and publicly display their own works. [16] These pre-existing regulations have already been protecting the copyrights of actors’ existing performances. Evan Sommer suggests that the reason copyright law has not been expanded to protect words generated by AI models is mainly that “human authorship is a bedrock requirement for a work to be copyrightable.” [17] Governments will therefore not register works made by AI or recognize them as original and creative content. Despite the lack of AI-specific restrictions on actors’ copyright, pre-existing regulations such as the Copyright Act already serve as a protective mechanism against all forms of copyright threats, which simultaneously includes protection from AI generative models.

IV. Legal Gaps in Existing Rights 
Although current legislation has been adapting to new changes in AI technology, existing laws and regulations are still insufficient in providing full protection for actors working in the industry. Since the Screen Actors Guild–American Federation of Television and Radio Artists’s 2024 strike, there have been new protections emphasizing transparency and informed consent from actors, particularly concerning AI-generated replicas. However, this policy mainly benefits those with the most negotiating power and networks, as it depends heavily on whether the actor can identify the unauthorized use of their identity or performance and negotiate agreements to control how it is used. [18] Instead of making companies collectively responsible for protecting actors, the system puts much of the responsibility on individual actors to protect themselves through contracts, creating conditions that breed vagueness and loopholes in the definitions and restrictions written into deals. Similarly, in the UK, performers’ rights are not fully protected under the Copyright, Designs and Patents Act 1988. UK intellectual property statutes only protect the recording of a performance, while performers are excluded from contractual agreements that credit or license the synthesis of their work. [19] This prevents actors from achieving appropriate financial gains, and with the emerging capabilities of generative AI models, it provides a loophole for the models to recreate and specifically exploit actors without violating the copyright laws. As illustrated by Ashley Beloat’s experience, the recreation of an actor’s performance through AI demonstrates how existing legal protections may struggle to address unauthorized digital replication, leaving performers vulnerable even when their original work is protected.

More importantly, these legal gaps and inefficiencies in previous amendments are mainly caused by the complexity of AI tech. Sommer suggests that the inaction of state governments and federal laws is mainly due to the complexity of AI technology, its rapidly advancing pace, and a lack of consensus on the best ways to approach regulating AI. [20] For instance, as AI generative models are only capable of collecting information from large databases rather than providing original and creative ideas, coincidences are much more likely to arise where a generated digital feature may have an individual likeness. An AI-generated actor may look different from the original actor but still hold recognizable aspects of their voice, movements, expressions, or performance. Conversely, if an AI system copies protected footage or other copyrighted material, copyright questions may arise alongside publicity-rights and contractual issues. These legal frameworks address different interests, and there is no one best solution to fully resolve every aspect of AI-generated replicas or emerging functions, considering its fast-paced developments.

V. Balancing Actors’ Rights with Creative Innovation 
To establish more holistic intellectual property rights and a morally sound working environment, companies should establish clearer disclosure and compensation standards without pushing responsibilities to the actors. Disclosure not only includes written informed consent on how digital replicas will be created and used, but it should also allow actors to gain active agency over their profile. This means that actors should be capable of tracking their own profile information even post-production, preventing scenarios where they only later discover their AI-generated replacements after a new series is released. At the same time, agreements should clarify whether actors receive additional payments when their digital replicas are used for new performances or commercial projects beyond the original production. This helps establish fair compensation for actors and provides avenues for them to protect their rights when their performances are used beyond the scope of the original agreement. More importantly, companies should uphold responsibilities to take control over the monitoring processes and become accountable for actors’ profile safety and maintenance. Transparency and confidentiality ultimately remain crucial in harnessing a safe legal environment for actors that respects their work and contributions.

Nevertheless, stronger protections should not prohibit all AI-generated content; they should distinguish authorized creative uses from exploitative or unauthorized uses of performers’ identities. Producers may have legitimate reasons in dubbing, visual effects, or localization, such as creating smoother transitions at a lower cost. Overly strict protections may hinder the beneficial side of these technologies, requiring more unnecessary and time-consuming work by producers. In such cases, technologies should instead spark and assist creative innovations. For instance, in Thaler v. Perlmutter (2025), the court ruled that a work created entirely by AI without human involvement could not qualify for copyright under the law as it stood [21]. Here, the ruling doesn’t restrict creative innovation; it provides a framework for ethical practices that prevents wrongful exploitation or copyright issues. This balance between assisting creative innovation and maintaining actors’ rights is crucial in the long-term development of the entertainment industry, allowing both workforce efficiency and humanistic care to be achieved.

VI. Conclusion
In conclusion, while existing U.S. publicity, copyright, and contract laws have laid important frameworks for bolstering actors’ intellectual property rights in the age of AI, the complexity and rapid developments of AI generative models continue to render these frameworks insufficient. As in the case of Ashley Beloat, the ability to create a performer’s recognizable identity raises questions that extend beyond ownership of an original production, challenging the boundaries of consent, authorization, and accountability. Addressing these challenges requires more than adapting existing laws to new technological developments. Proactive measures should be taken to prevent AI violations rather than following the generative model’s footsteps. Ultimately, the future of AI in entertainment should not be measured solely by production efficiency, but also by whether the industry honors the rights and contributions of the people whose identities make its content so valuable.

[1] Lucia Moses, Actors are losing roles to AI knockoffs of themselves, Business Insider, 2026, https://www.businessinsider.com/actors-see-ai-versions-of-their-performances-in-new-shows-20 26-7 (last visited Oct. 16, 2026). 

[2] Sarah Thomas, AI and Actors: Ethical Challenges, Cultural Narratives and Industry Pathways in Synthetic Media Performance, 2 EMERGING MEDIA 523 (2024). [3] Ibid. 

[4] Matthew Tan, China’s short-drama producers flood the market with cheap bets - and let audiences pick the winners, CNBC, 2026, 

https://www.cnbc.com/2026/08/26/short-drama-china-production-ai-entertainment-economics.ht ml (last visited Oct. 16, 2026). 

[5] Lin Ying-Hsuan, Short-form video, ai-generated dramas power next wave of global content as filmart panel probes the thriving Digital economy, Variety, 2026, 

https://variety.com/2026/digital/news/short-form-video-ai-generated-dramas-filmart-1236692132 / (last visited Oct. 16, 2026). 

[6] Toby Butterfield and Liberty McAteer Authored an Article Titled, “Using AI-Generated Actors and Models is Creating New Legal Issues.,” Moses & Singer LLP, 2026, https://www.mosessinger.com/publications/toby-butterfield-and-liberty-mcateer-authored-an-arti cle-titled-using-ai-generated-actors-and-models-is-creating-new-legal-issues (last visited Oct 16, 2026). 

[7] Eric Sefton, Practice tips: Protecting celebrity likenesses in the age of generative AI, Los Angeles County Bar Association, 2026, 

https://lacba.org/?pg=LosAngelesLawyerMagazine&pubAction=viewIssue&pubIssueID=67427 &pubIssueItemID=448444 (last visited Oct. 16, 2026). 

[8] Ibid. 

[9] Assemb. Bill 1836, 2023-2024 Reg. Sess., ch. 258, 2024 Cal. Stat. 

[10] Ibid. 

[11] N.Y. Civ. Rights Law §§ 50, 51 (McKinney 2026). 

[12] Assemb. Bill 2602, 2023-2024 Reg. Sess., ch. 257, 2024 Cal. Stat. 

[13] Ibid. 

[14] N.Y. Lab. Law § 1035(10) (McKinney 2026).

[15] Industry Standards & Legal Developments for the AI Era, Rodriques Law: Entertainment Blog, 2026, https://rodriqueslaw.com/entertainment/ai-industry-standards. [16] 17 U.S.C. §§ 102, 106 (2026). 

[17] Evan Sommer, “Real Concerns for an Artificial Threat: Artists, AI, and the Battle to Script Hollywood’s Future,” 25 NEV. L.J. 449 (2025). 

[18] Supra note 2. 

[19] Ibid. 

[20] Ibid. 

[21] Ibid.