Fresh off scaring the bejeezus out of many in Hollywood with demos of its text-to-video generator Sora, OpenAI now wants in. According to Bloomberg, top executives at the generative AI developer will hold a round of meetings this week with a number of film studios and Hollywood honchos to discuss what Sora can do for them.
We’ve discussed here before why everything that can be made with AI, will be in Hollywood. So it is no great surprise that studio folks would take the meetings. But appearing to get cozy with Sora right now carries significant risk for the studios.
Generative AI was recently at the center of extensive labor unrest in Hollywood that cost the studios the better part of a year’s worth of production. As a result of that unrest, they are also now bound by collective bargaining agreements with writers and actors that circumscribe what they can do unilaterally with tools like Sora.
They also face the possibility of further AI-related disruption when the current contract with the Hollywood craft unions expires at the end of July.
Then there’s the tricky question of where Sora came from. OpenAI is facing a passel of copyright infringement lawsuits against it over its use of copyrighted works in training its models, including a multi-billion dollar claim from the New York Times. Insofar as Sora may have been trained on copyrighted videos, it could be a bad look for the studios, and the media conglomerates that own them, to be jumping in bed too quickly with an accused infringer.
In a recent interview with the Wall Street Journal, OpenAI CTO Mira Murati claimed, rather implausibly, not to know — or, less charitably, declined to reveal — whether Sora had been trained on videos scraped from YouTube and other social media platforms.
That likely had something to do with fear of attracting additional litigation. But we may find out anyway. Under the European Union’s AI Act, now poised to go into force, OpenAI could be required to provide a “sufficiently detailed summary” of the datasets used in training its models, which could prove embarrassing to any studio making extensive use of Sora.
In addition, the U.S. Copyright Office is preparing to release a series of reports this year addressing various issues around AI and copyright, along with recommendations to Congress regarding possible changes to copyright law. Among the topics to be addressed is the question of whether works created using AI tools are wholly eligible for copyright protection. If the Office limits eligibility in a way that renders AI generated elements of a movie unprotectable, it could make over-reliance on AI tools in production too risky.
All those risks are for the future, however. For now, the studios are desperate to cut costs. And that immediate need is likely to override any long-term risks.