Last month, clips from an interview with OpenAI's Chief Technology Officer Mira Murati went viral when The Wall Street Journal asked the executive what data OpenAI's video generator Sora uses—and whether it pulled from YouTube. "I'm actually not sure about that," Murati claimed of whether Sora was trained on YouTube video data.

Now, YouTube CEO Neal Mohan tells Bloomberg that if OpenAI used his firm's videos, that would be a problem.

"From a creator’s perspective, when a creator uploads their hard work to our platform, they have certain expectations," Mohan states. "One of those expectations is that the terms of service is going to be abided by. It does not allow for things like transcripts or video bits to be downloaded, and that is a clear violation of our terms of service. Those are the rules of the road in terms of content on our platform," the CEO continued.

Strangely, neither Murati nor Mohan are willing to take a clear stance on whether Sora was actually trained on YouTube's data. "I have seen reports that it may or may not have been used. I have no information myself," said Mohan, who urged Bloomberg to ask OpenAI.

Recommended by Our Editors

Google's Agentic AI Tool Gemini Spark Is Now Available. Here's How to Try It

I Used ChatGPT to Build Complex Spreadsheets Fast. Here’s How You Can, Too

GitHub Copilot Costs Skyrocket As Users Are Pushed to Per-Token Billing

But OpenAI has already dodged the question, potentially because of its ongoing legal battles over alleged copyright violations. Previously, Murati gave only a vague, general answer about Sora's training data, offering that OpenAI used "publicly available data and licensed data" to get Sora to generate videos from text prompts.

Murati wouldn't even confirm whether Sora was trained on Shutterstock videos—a company OpenAI signed a six-year training data agreement with last year. Shutterstock's release states that OpenAI has access to its images, videos, and music libraries for training purposes.

Not all YouTube videos are entirely off-limits, however. Mohan pointed out that some creators have struck specific licensing deals of their own, and Google may be using "some portion" of YouTube's content trove to train its own Gemini models. Mohan says YouTube is following its own terms of service and abiding by any existing contracts for its own AI model training.

Since announcing Sora in February, OpenAI has shown off its AI tool to executives at Paramount, Warner Bros., and Universal in hopes that Hollywood might use it in some capacity. Sora is expected to launch sometime later this year.