Stay informed with free updates
Simply sign up to the US & Canadian companies myFT Digest — delivered directly to your inbox.
OpenAI’s new video-generation artificial intelligence tool Sora was briefly leaked by artists testing the new model, which some creative industries fear could pose a disruptive threat.
A project on developer platform Hugging Face on Tuesday appeared to be connected to the application programming interface for OpenAI’s Sora, enabling others to access the model and generate videos using the AI.
Several testers uploaded a letter to the platform accusing OpenAI of taking advantage of “hundreds of artists [who] provide unpaid labour through bug testing, feedback and experimental work” in so-called “redteaming” of Sora — a term used for early testing of AI models. After several hours, OpenAI shut off early access for artists, according to a message on the Hugging Face page.
It illustrates the conflict between the start-up behind the leading AI model, recently valued at $150bn, and the artists who stand to lose the most from Sora, which can turn text into video. OpenAI’s chief executive Sam Altman and other executives have this year been courting Hollywood executives, to showcase Sora and assuage fears it may lead to a loss of jobs in films.
Dubbing themselves “PR Puppets”, the letter was addressed to “corporate AI overlords” and accused OpenAI of “artwashing”.
“We are not against the use of AI technology as a tool for the arts . . . What we don’t agree with is how this artist programme has been rolled out and how the tool is shaping up ahead of a possible public release.”
Sora was unveiled at the beginning of this year but has not been widely released. OpenAI has only granted access to a small number of visual artists, designers and filmmakers to gather feedback on its potential harms, risks and areas for improvement.
According to the artists, only minimal compensation was offered to testers, while some would be able to compete to have films created with Sora screened.
“This early access programme appears to be less about creative expression and critique, and more about PR and advertisement,” it added.
The collective of artists and signatories confirmed the letter to the Financial Times. TechCrunch first reported the leak.
OpenAI said Sora was still in preview mode for research purposes, and it was “working to balance creativity with robust safety measures for broader use”.
It added participation in its testing was voluntary with “no obligation to provide feedback” — and artists using the tool are required to refrain from sharing confidential details while it is under development.
Last month, more than 11,000 creative professionals — including actors Julianne Moore and Kevin Bacon, Radiohead singer Thom Yorke and a host of writers and musicians — signed an open letter that called generative AI “a major, unjust threat” to creatives’ livelihoods.
In a recent discussion on Reddit, OpenAI’s chief product officer Kevin Weil said Sora’s wider release had been delayed by the “need to perfect the model”, including in areas such as safety and scaling computing power for the model.
Read the full article here