YouTube officially launched a new feature this Monday, granting creators direct control over how third-party companies utilize their video content to train artificial intelligence models. This update arrives as a direct response to growing concerns regarding unauthorized data scraping and the use of creator material without consent or compensation.
Managing AI Training Permissions
Through the YouTube Studio dashboard, eligible creators and rights holders—specifically those with administrator access—can now opt in to authorize specific third-party entities to train AI models on their videos. A new settings interface displays a list of 18 initial partners, including major industry players such as OpenAI, Meta, Microsoft, Apple, Nvidia, Anthropic, and Amazon, among others.
In addition to selecting individual companies, creators have the option to enable a “All third-party companies” setting, granting broad permission for any AI developer to utilize their data. By default, the system is set to exclude all third parties, reinforcing YouTube’s stance that previous unauthorized training by external entities was contrary to its Terms of Service.
Scope and Future Implications
While this feature offers significant control, it does not alter Google’s own internal practices. The company confirmed that it will continue to train its proprietary AI models on YouTube content, adhering to existing agreements with creators. Furthermore, the new setting does not change the platform’s underlying policy, which continues to prohibit unauthorized scraping.
YouTube views this rollout as the inaugural step toward a more transparent ecosystem. Future updates are expected to streamline the process, potentially allowing authorized companies to access direct video downloads, which could pave the way for formal compensation models for creators.
Broader AI Strategy
This initiative follows YouTube’s September commitment to address AI-related concerns. Beyond training controls, the platform is expanding its Content ID system to include detection tools designed to protect the likeness, voices, and faces of creators, musicians, and athletes from unauthorized AI generation.
As the industry evolves, YouTube continues to balance its ecosystem with the rapid pace of AI development. Notably, this announcement coincides with Google DeepMind’s reveal of Veo 2, a new video-generating AI model positioned to compete directly with OpenAI’s Sora. Creators globally will receive notifications regarding these new dashboard controls via desktop and mobile banners in the coming days.