YouTube announced on Tuesday that it is extending access to its likeness detection technology—designed to identify AI-generated deepfakes—to a pilot group of government officials, political candidates, and journalists. Participants in this pilot will be able to use a tool that spots unauthorized AI-generated content and submit requests for its removal if they believe it breaches YouTube’s policies.

This technology was first introduced last year to approximately 4 million creators in the YouTube Partner Program, after earlier testing phases. Much like YouTube’s Content ID system, which scans for copyrighted material in uploaded videos, the likeness detection feature looks for AI-simulated faces. Such tools are sometimes employed to spread misinformation by using the fabricated personas of prominent individuals—such as politicians or officials—to show them saying or doing things they never actually did.

Through this new pilot, YouTube seeks to balance user free expression with the risks posed by AI that can produce convincing replicas of public figures. Leslie Miller, YouTube’s Vice President of Government Affairs and Public Policy, emphasized in a press briefing ahead of Tuesday’s launch, “This expansion is really about the integrity of the public conversation.” She added, “We know that the risks of AI impersonation are particularly high for those in the civic space. But while we are providing this new shield, we’re also being careful about how we use it.”

Miller clarified that not every detected match will be automatically removed upon request. Instead, YouTube will assess each case under its existing privacy guidelines to determine whether the content constitutes protected expression, such as parody or political critique. The company also noted it is advocating for federal protections, including support for the NO FAKES Act in Washington, D.C., which would regulate the use of AI to create unauthorized recreations of a person’s voice and visual likeness.

To use the tool, eligible pilot testers must first verify their identity by uploading a selfie and a government ID. They can then create a profile, review matches that appear, and optionally request removal. YouTube stated it eventually plans to allow people to block violating content before it goes live or, potentially, to monetize those videos—similar to how its Content ID system operates. The company did not specify which politicians or officials will be part of the initial testing but indicated the goal is to broaden access to the technology over time.

AI-generated videos will be labeled, though label placement is not uniform. For some videos, the label appears in the description, while content dealing with more “sensitive topics” will have the label displayed prominently on the video itself. This mirrors YouTube’s approach to all AI-generated content. Amjad Hanif, YouTube’s Vice President of Creator Products, explained the reasoning: “There’s a lot of content that’s produced with AI, but that distinction’s actually not material to the content itself. It could be a cartoon that is generated with AI. And so I think there’s a judgment on whether it’s a category that maybe merits from a very visible disclaimer.”

YouTube has not disclosed how many AI deepfake removals have been processed since creators gained access to the detection tool, but noted the volume taken down so far has been “very small.” Hanif remarked, “I think for a lot of [creators], it’s just been the awareness of what’s being created, but the volume of actually removal requests is really, really low because most of it turns out to be fairly benign or additive to their overall business.” However, that may not hold true for deepfakes targeting government officials, politicians, or journalists. Looking ahead, YouTube intends to expand its deepfake detection technology to cover additional areas, including recognizable voices and other intellectual property like popular characters.