YouTube is moving its AI likeness detection work into a broader testing phase, bringing a small group of major creators into a pilot meant to identify and manage synthetic content that imitates real people.
The company announced the expansion on Wednesday alongside public support for the NO FAKES ACT, legislation aimed at AI-generated replicas that simulate a person’s image or voice in misleading or harmful content.
What YouTube Is Testing
The pilot focuses on AI-generated content that features the “likeness” of creators, artists, and other famous or influential figures. In the source article, that likeness includes a person’s face, and YouTube has also described simulated voices as part of the problem it is trying to address.
The technology was introduced in partnership with the Creative Artists Agency (CAA) in December 2024. It builds on YouTube’s existing Content ID system, which detects copyright-protected material in uploaded videos.
The comparison to Content ID matters because it shows how YouTube is framing the new tool. The goal is not only to respond after a person finds a fake video, but to automatically detect content that may violate rules around AI-generated depictions.
The First Named Pilot Testers
YouTube is now naming the initial participants in the pilot. The list includes some of the platform’s most visible creators and channels:
- MrBeast
- Mark Rober
- Doctor Mike
- the Flow Podcast
- Marques Brownlee
- Estude Matemática
During the testing period, YouTube plans to work with these creators to scale the technology and improve its controls. The company also said the program will expand to more creators over the year ahead.
One important detail remains open: YouTube did not say when the likeness detection system will become more publicly available. For now, the expansion is still a controlled pilot rather than a platform-wide launch.
Why the NO FAKES ACT Is Part of the Story
Alongside the product update, YouTube said it supports the NO FAKES ACT. The company says it collaborated on the bill with Sens. Chris Coons (D-DE) and Marsha Blackburn (R-TN), as well as industry groups including the Recording Industry Association of America (RIAA) and the Motion Picture Association (MPA).
Coons and Blackburn were set to announce the reintroduction of the legislation at a press conference on Wednesday. YouTube’s position is that platforms need a clearer way to tell the difference between authorized AI content and harmful fakes.
“We also know there are risks with AI-generated content, including the potential for misuse or to create harmful content. Platforms have a responsibility to address these challenges proactively,”
That statement captures the tension YouTube is trying to address. The company says AI could “revolutionize creative expression,” but it also sees a need for systems that give people a way to challenge AI-generated depictions of themselves.
How YouTube Wants the Process to Work
YouTube’s argument for the NO FAKES ACT centers on notification. The company says individuals should have the ability to notify platforms about AI-generated likenesses they believe should be removed.
According to YouTube, that process is necessary because platforms need input from the people being depicted. Without that signal, the company argues, platforms cannot make informed decisions about whether content is authorized or harmful.
This is also where the likeness detection pilot fits into YouTube’s broader approach. Detection can help find possible violations, while controls and notification tools can help decide what happens next.
Existing Tools Around Synthetic Likeness
The pilot is not YouTube’s only move in this area. The company previously updated its privacy process so people can request removal of altered or synthetic content that simulates their likeness.
It also added likeness management tools that allow people to detect and manage how AI is used to depict them on YouTube. Together, those measures suggest a layered approach: removal requests, management tools, and automated detection working around the same core issue.
For creators, the practical question is control. AI can make it easier to copy a face or voice, and YouTube is testing systems that could help creators find those uses and respond to them. For the wider platform, the challenge is separating authorized creative uses from AI fakes that mislead viewers or cause harm.