YouTube is introducing a feature that allows creators to voluntarily let third-party companies use their videos to train AI models. The option will be off by default, meaning creators must actively enable it if they wish to participate. YouTube notes in another post that third-party companies may not end up using the content after all, even if creators allow them to, and that YouTube itself can't control what a third-party company does with their content.
According to a support post by TeamYouTube, the platform sees this as a way for creators to explore new opportunities in the AI age. “We see this as an important first step in supporting creators and helping them realise new value for their YouTube content in the AI era,” a YouTube representative named Rob stated. "As we gather feedback, we'll continue to explore features that facilitate new forms of collaboration between creators and third-party companies, including options for authorised methods to access content."
YouTube notes that if creators change their mind and want to opt out, they can turn off the setting at any time or change which third-party companies they allow to use their videos for AI training. The initial list, first published by TechCrunch and then confirmed by YouTube spokesperson Jack Malon to The Verge, includes major players in AI development such as OpenAI, Meta, Microsoft, Adobe, Nvidia, and Apple.
The new setting will appear in YouTube Studio over the next few days, so that creators can select specific companies from a provided list or permit all approved companies to use their videos.
These companies were chosen based on their active work in generative AI, Malon explained, adding that they are likely candidates for meaningful partnerships with creators. Unauthorised scraping of content for AI training, however, remains against YouTube’s policies.
This announcement comes after reports of large AI firms, including OpenAI and Anthropic, using content scraped from YouTube to train their models.
