Artificial intelligence developer OpenAI is partnering with major Hollywood talent agencies and actor Bryan Cranston to combat the proliferation of AI-generated deepfakes, marking a significant policy shift after widespread concerns over unauthorized content.
The alliance, announced Monday, involves the Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA), United Talent Agency (UTA), Creative Artists Agency (CAA), and the Association of Talent Agents (ATA). It aims to prevent the unauthorized use of actors’ identities.
This move follows public outcry and specific incidents, including deepfake videos featuring Cranston’s likeness and voice. These clips appeared online shortly after the release of OpenAI’s Sora 2 video tool in late September.
Cranston, known for his role in “Breaking Bad,” stated he was unaware of these viral clips until they gained traction, sparking indignation within the artistic community.
Through a statement issued by SAG-AFTRA on X, Cranston expressed his gratitude for OpenAI’s updated policies. “I am grateful to OpenAI for their policy and for improving their protection mechanisms, and I hope they and all companies involved respect our personal and professional right to control the replication of our voice and image,” he said.
OpenAI CEO Sam Altman formalized a key policy change on October 3. The company will no longer assume it can use artists’ materials unless the rights holder explicitly objects.
This new approach grants actors and their agencies “granular control over the generation of characters.” It gives them direct power over how and when their identities can be replicated by AI.
Altman also reaffirmed OpenAI’s support for the NO FAKES Act, a proposed U.S. law designed to curb unauthorized AI replicas. “OpenAI is deeply committed to protecting performers from misappropriation of their voice and image,” Altman stated. “We were an early supporter of the NO FAKES Act when it was introduced last year, and we will always defend artists’ rights.”
This alliance comes after months of criticism toward OpenAI from talent agencies. These agencies accused the company of using copyrighted material to train its AI models. Both CAA and UTA had previously labeled Sora a “direct threat” to their clients’ intellectual property.
Past controversies included users uploading “disrespectful” videos of Martin Luther King Jr. generated with Sora, which OpenAI removed upon request from his estate. Zelda Williams, daughter of the late Robin Williams, also publicly appealed for users to stop sending her AI-generated videos mimicking her father.
OpenAI had already required explicit consent for using a person’s voice or image. However, under pressure, the company has now pledged to respond more quickly to complaints regarding impersonations or misuse.
The company, which has invested billions of dollars into data centers and AI chips, seeks to avoid lawsuits and restore confidence. This includes rebuilding trust with actors, agencies, and lawmakers.
This cooperation symbolizes a crucial step in the evolving relationship between Silicon Valley and the entertainment industry. It seeks to redefine the boundaries between human creativity and digital automation.