Hollywood agencies accuse OpenAI of misleading them about Sora 2 protections

Hollywood talent agencies are accusing OpenAI of deliberately misleading them about Sora 2’s content protections after the AI video generator launched with capabilities to create clips featuring copyrighted characters and movie scenes. The controversy deepens existing tensions between the AI industry and entertainment sector over intellectual property rights and unauthorized use of creative content.

What you should know: Major talent agencies claim OpenAI either failed to notify them of Sora 2’s launch or was “purposefully misleading” about the strength of its content guardrails.

• WME, a major Hollywood talent agency that represents Ben Affleck, Christian Bale, Matt Damon, Denzel Washington, and Jennifer Garner, argues OpenAI should require permission from celebrities, studios, and animators before allowing their characters in AI-generated videos.
• Pre-launch meetings with OpenAI executives, including COO Brad Lightcap, were described as “upbeat,” with the company appearing confident about protecting intellectual property and client likeness.

Key loopholes exploited: Despite launching with a ban on creating videos of public figures, copyrighted content slipped through Sora 2’s safeguards.

• Users readily created scenes from recognizable properties including Bob’s Burgers, SpongeBob SquarePants, Gravity Falls, Pokémon, Grand Theft Auto, and Red Dead Redemption.
• The tool enabled mashups of well-known titles, such as making Pokémon’s Pikachu appear as a character in The Lord of the Rings and Oppenheimer.
• Users discovered an unexpected loophole for creating videos of dead celebrities like Michael Jackson, which OpenAI banned a week later after clips went viral.

Industry backlash: Hollywood’s response was swift and decisive following Sora 2’s public release.

• WME released a memo on launch day opting out all clients from AI-generated videos.
• CAA, another major talent agency representing Scarlett Johansson and Tom Hanks, called the product “exploitation.”
• The controversy follows Johansson’s 2024 dispute with OpenAI over allegedly copying her voice.

What they’re saying: OpenAI CEO Sam Altman admitted the product’s main goal was financial and entertainment value.

• “This was a very calculated set of moves [CEO Sam Altman] made,” a WME executive told The Hollywood Reporter, suggesting OpenAI “knew exactly what they were doing when they released this without protections and guardrails.”
• Altman claims he’s hearing excitement from content creators about “this new kind of ‘interactive fan fiction,'” according to an October 4 blog post.

The big picture: Sora 2’s viral clips propelled it to the top of app stores, though its Apple Store ratings have since fallen to 2.8.

• OpenAI promises to give “rightsholders more granular control over generation of characters, similar to the opt-in model for likeness but with additional controls,” but hasn’t disclosed specific solutions or timelines.
• WME is exploring litigation against OpenAI, though legal guidelines for AI-generated content remain evolving.
• The entertainment industry scored a recent victory when a judge ordered Anthropic, an AI company, to pay $1.5 billion for training models on copyrighted books.