Youtube expands deepfake detection to politicians and journalists, allowing them to verify unauthorized image use

YouTube has expanded its deepfake detection tool to include politicians and journalists, enabling them to check if their images are being used in unauthorized videos.

Tool to combat deepfakes

The platform initially launched this feature in October 2024 as a way to help content creators fight deepfakes that use their image and voice in videos without their consent. The tool allows creators to monitor how their image appears, created or modified with AI, to protect their identity. It's especially useful for manipulated videos that reflect a viewpoint they don't share.

Now, YouTube has extended this tool to government officials, politicians, and journalists as part of a pilot test, so they can ensure their images aren't being used in deepfake videos without permission.

According to the company's blog post, this tool functions similarly to Content ID, YouTube's system for detecting and protecting copyrighted content. In this case, it focuses on image detection and searching for faces of people enrolled in the deepfake detection program.

If this system finds a match, the person in the program can request the content be removed from the platform, though this requires YouTube approval based on its privacy policies.

This offers an effective way to identify potential unauthorized substitutions, but doesn't guarantee removal. This is because, for public figures, YouTube protects free expression by preserving content like parody and satire, even when used to criticize world leaders or influential figures.

To participate in the deepfake detection program, politicians and journalists must verify their identity by sharing a video of themselves and an official ID. These data will be used exclusively for identity verification, not to train generative AI models at Google.