The recent announcement of YouTube’s new AI detection tools marks a significant shift in how creators – actors, musicians, and digital content creators – manage their intellectual property. Earlier this month, YouTube announced that it is developing new tools to prevent the copying, altering, and general misuse of creators’ likeness by unauthorized third parties on the platform. These tools include: (1) voice identification technology that will “allow partners to automatically detect and manage AI-generated content on YouTube that simulates their voices,” and (2) AI generated face detection technology that detects and manages AI-generated content showing creators’ faces on the site.
The voice identification technology seems to be the long-awaited answer to a pledge made last year by YouTube to come up with a way to compensate artists whose work was used to create AI music. Since then, YouTube has collaborated with Universal Music Group (UMG)’s talent to help craft a solution. According to the latest announcement from YouTube, this new technology should not only be able to detect the content simulating their voices but should also easily identify which rightsholders should be paid when their works are used in AI music.
The AI generated face detection technology is a potential solution many creators are eager to see in practice as misleading endorsements, advertisements, and general misinformation containing creators’ likeness has become more prevalent with the rise of AI tools. Although this solution is in the earlier stages of development, the platform has divulged a few details regarding the takedown process, including that rightsholders will be given a menu of options from takedowns to splitting ad revenue.
These advancements build off of ContentID’s existing digital fingerprinting system that has made it easy for partners to identify and manage copyrighted content on YouTube since 2007. ContentID has been instrumental in enabling partners to monetize their content through royalty payments, processing billions of claims and generating billions of dollars in revenue each year for creators and artists. The newest proposed additions to the ContentID platform will likely make it much easier for rightsholders to assert claims in the age of AI.
It is important to note that, in announcing these updates, YouTube reiterated that the unauthorized use of YouTube videos to train AI learning models violates its Terms of Service. As the debate over the legality of using copyrighted content to train AI without permission continues, developments like this will continue to shape how creators can protect their IP on social platforms. Staying abreast of these legal and technical changes will be key as courts and tech giants alike navigate the next steps of the AI revolution.
If you need assistance in the areas of AI, digital media, and/or copyright infringement, please contact our attorneys here.
