The company is giving ‘big name’ directors and actors access to its AI video generator — but still won’t say what it’s trained on.
What we know about OpenAI’s plan for Sora in tinseltown. Credit: SOPA Images / Getty Images
OpenAI is ramping up plans for its AI video generator Sora — and that involves a charm offensive in Hollywood.
Details are still fuzzy, but we know the company is approaching filmmakers as well as studios. OpenAI CEO Sam Altman and COO Brad Lightcap are having “introductory conversations” with industry stakeholders, according to Bloomberg.
Unspecified “big name” directors and actors already have access to Sora, the report says. That’s part as an effort to “encourage filmmakers to integrate its new AI video generator into their work.”
Sora was unveiled in February — and though there hasn’t been a public release yet, the announcement has raised concerns about the data that was used to train to model, and how it could impact the film industry.
As with ChatGPT, OpenAI hasn’t been transparent about Sora’s training data. But creatives already suspect Sora was trained by scraping art and videos without the knowledge or consent of their creators.
OpenAI is already facing several copyright infringement lawsuits including allegations of this practice with the large language models that power ChatGPT.
The use of AI video tools threatens to upend the film industry by replacing jobs that range from VFX professionals, to writers, and even to actors.
The recent strike by Hollywood writers’ and actors’ unions (WGA and SAG-AFTRA) strike sought contractual limits on the use of AI in writers’ rooms. The unions also fought to create digital likenesses of actors that could be used in perpetuity without pay.
The WGA is voting soon on a tentative agreement that prevents AI content being used as source material for writers’ rooms. SAG-AFTRA, in its contract with studios, won promises of compensation and credit for AI likenesses — but did not succeed in banning the practice altogether.
Already, images featuring generative AI have crept into films such as Late Night with the Devil.
‘Artistwashing’
Meanwhile, OpenAI published a blog post full of “first impressions” from a select group of testers who are visual artists, filmmakers, and creative directors — which tells a different story.
The post features feedback from director Paul Trillo, production company shy kids, creative agency Native Foreign, artist August Kamp, creative director Josephine Miller, and AR/XR artist Don Allen Stevenson III. Unsurprisingly, the feedback published on OpenAI’s blog was overwhelmingly positive.
Testers praised Sora’s ability to create photorealistic videos from text prompts and without constraints. “Not restricted by time, money, other people’s permission, I can ideate and experiment in bold and exciting ways,” said Trillo.
But users on X were quick to point out OpenAI’s controlled narrative. “Artistwashing: when you solicit positive comments about your generative AI model from a handful of creators, while training on people’s work without permission/payment,” wrote Ed Newton-Rex, CEO of ethical AI data sourcing non-profit Fairly Trained.
If OpenAI is gearing up to take on Hollywood, in other words, the company had better be prepared for cinematic drama.
Cecily is a tech reporter at Mashable who covers AI, Apple, and emerging tech trends. Before getting her master’s degree at Columbia Journalism School, she spent several years working with startups and social impact businesses for Unreasonable Group and B Lab. Before that, she co-founded a startup consulting business for emerging entrepreneurial hubs in South America, Europe, and Asia. You can find her on Twitter at @cecily_mauran.
This newsletter may contain advertising, deals, or affiliate links. Subscribing to a newsletter indicates your consent to our Terms of Use and Privacy Policy. You may unsubscribe from the newsletters at any time.
