Authors
Introduction
For decades, films imagined AI as a futuristic threat. Today, it has become a practical production tool that is actively transforming how film and television content is created, edited, and distributed. Rather than replacing creative professionals wholesale, to date, AI is increasingly being used to accelerate workflows, enhance visual effects, improve localization, and support creative decision-making. At the same time, its rapid adoption has prompted fundamental legal and commercial concerns around copyright, performers’ rights, and ownership of creative works.
One of the most visible applications of AI technology remains digital de-ageing and digital performance enhancement. Following The Irishman (2019) and Indiana Jones and the Dial of Destiny (2023), AI-assisted techniques have become increasingly sophisticated and are now routinely used to recreate younger versions of performers, restore archival footage, and generate seamless visual effects. AI-powered dubbing, vubbing and voice adaptation have also gained traction, enabling productions to create more natural multilingual versions while preserving the original actor’s performance. In 2024’s The Brutalist, film editor Dávid Jancsó revealed that the production used voice-generating technology to enhance the actors’ Hungarian dialect. Studios are increasingly exploring AI-assisted localization, automated subtitle generation, and culturally tailored edits for global audiences.
Behind the camera, AI has become an important production tool. Studios and production companies are using generative AI to assist with script analysis, budgeting, scheduling, storyboarding, concept art, virtual production, editing, visual effects, marketing asset generation, and archive management. While these technologies can significantly reduce time and cost, they remain subject to increasing contractual and legal scrutiny.
The industry’s embrace of AI has also accelerated commercial partnerships. In 2024, Creative Artists Agency expanded its relationship with Metaphysic, launching a dedicated business to help clients commercialize and protect their digital likenesses. The partnership reflects a broader shift away from viewing digital replicas as purely visual effects tools, toward treating them as valuable intellectual property that can be licensed, managed, and monetized throughout an artist’s career and potentially beyond.
Unions and performers’ rights
Perhaps the most significant AI-related developments in the film and TV sector have come through collective bargaining. Following the 2023 Hollywood strikes, both the Writers Guild of America (WGA) and the Screen Actors Guild - American Federation of Television and Radio Artists (SAG-AFTRA) secured landmark contractual protections governing generative AI in the context of individual performers and writers.
The WGA agreement confirms that AI-generated material is not “literary material” and cannot itself receive writing credit. Writers may choose to use AI as a creative tool, but companies cannot require them to do so, nor can AI-generated content undermine writers’ authorship or compensation. The agreement also provides safeguards related to the disclosure of AI-generated materials and the use of writers’ work to train AI systems, signaling the industry’s determination that AI should complement rather than replace human creativity, and that control over its use should remain with the talent.
SAG-AFTRA’s agreement goes further in regulating the use of digital replicas. Producers must generally obtain informed consent before creating or using digital replicas of performers and must negotiate appropriate compensation. The agreement distinguishes between different categories of digital replicas, including those created during production and those generated independently using AI. These provisions reflect growing concern over voice cloning, facial replication, and the creation of synthetic performances that could otherwise be reused indefinitely without further involvement from performers and/or further compensation. SAG-AFTRA executive director Duncan Crabtree-Ireland noted that the union’s agreement will “ensure synthetics remain the exception in [the film and television] industry instead of the rule.”1
The Directors Guild of America has similarly recognized that AI should remain subject to human creative consultation and oversight, while other guilds and industry bodies continue to develop guidance on responsible AI adoption. Collectively, these agreements have established an emerging industry standard in which transparency, consent, and human authorship remain central principles.
These issues also extend beyond living performers. Digital replicas, synthetic voices, and posthumous performances have become increasingly valuable commercial assets. Estates are beginning to license the digital likenesses of deceased actors, while advances in voice synthesis have made it technically possible to recreate performances with remarkable realism. In 2026, it was announced that deceased actor Val Kilmer would star in As Deep as the Grave, a film he had originally been cast in before his death from cancer in 2025. As these technologies become more accessible, contractual provisions governing consent, scope of use, duration, and compensation are likely to become standard features of talent agreements.
Copyright, trademark, and AI training data
Alongside collective bargaining, one of the most significant legal issues in AI adoption and use in the film and TV industry concerns the datasets used to train generative AI models. A growing number of lawsuits in the United States have been brought by authors, publishers, artists, and media companies against AI developers, alleging that copyrighted works were used without permission to train large language models and image generation systems.
Although these cases do not focus exclusively on film and television, their outcomes could have significant implications for the screen industries. If courts conclude that AI developers require licenses to use copyrighted material for training, studios may increasingly favor AI tools trained on licensed datasets. Conversely, broader interpretations of fair use could accelerate adoption of more capable foundation models across creative industries.
These disputes also affect businesses developing proprietary AI systems using internal script libraries, production archives, or audiovisual catalogs. Rights holders are increasingly scrutinizing how historic content is being used to develop new AI capabilities, making data provenance and licensing an important emerging aspect of AI governance.
Individual performers are also taking steps to protect themselves outside of copyright law. Celebrity talent in the United States has sought protection over their names, likenesses, and personas, using trademark law as an additional line of defense against unauthorized AI-generated replicas of their voice or image. Actor Matthew McConaughey made headlines by securing trademark protection for his famous “Alright, alright, alright!” delivery, later explaining that he wanted to “create a clear perimeter around ownership with consent and attribution the norm in an AI world.”2 This trend suggests that alongside copyright litigation and collective bargaining, publicity and trademark rights may become an additional legal tool for talent seeking to control how their identity is used to train or generate AI content.
Further compliance issues arise in terms of privacy rules given such user cases involve the processing of personal data and in Europe, the General Data Protection Regulation will almost certainly apply. Resulting obligations of transparency and purpose limitation become complicated where archive footage is used for AI training which was not anticipated at the time of original collection. In addition, a lawful basis is needed for the processing and mistakes are routinely still made by controllers in this area including reliance on consent where this is not appropriate (even though “consent” in the broader sense may be obtained separately as a release but this is not the same thing under data protection laws) and failing to consider whether the processing involves special category data and how any data subject rights will apply.
Looking ahead
The debate has shifted from whether AI will be used in film and television to how it can be deployed responsibly. Increasingly, studios are implementing internal AI governance frameworks that require legal review, disclosure of AI-assisted workflows, and human oversight for creative decisions.
At the same time, technology companies are investing in provenance tools, watermarking, and content authentication to distinguish AI-generated material from authentic human-created content. While no technical standard has yet emerged, these technologies may become increasingly important in protecting intellectual property and combating misinformation.
We are also seeing a rise in dedicated media generation tools powered by AI, purporting to democratize the production industry and provide individuals with powerful tools to generate studio quality content and edits.
Commercial contracts are also evolving. AI-specific clauses covering disclosure, training data, ownership of outputs, digital likeness rights, and permitted uses are becoming significantly more common in production agreements, commissioning arrangements, and talent contracts.
Looking ahead, regulators and industry players are likely to continue focusing on synthetic performances and AI-generated effects and content. As legal frameworks mature and litigation clarifies the boundaries of copyright and training data, the industry’s current approach is likely to remain: AI should enhance human creativity, not displace it. Achieving that balance will require continued collaboration between technology companies, studios, performers, writers, and regulators as the next generation of creative tools becomes increasingly sophisticated.
1. au.variety.com2. bbc.com