YouTube is expanding its AI deepfake detection tool to include politicians and journalists, offering them the ability to monitor AI-generated deepfakes featuring their likeness on the platform. This tool, already accessible to numerous content creators on YouTube, will now be extended to a select group of journalists, government officials, and political figures. The new likeness detection feature operates akin to Content ID but focuses on identifying individuals’ faces within videos. If a match is found, those enrolled in the program can request YouTube to take down the content, although not all requests will be granted, considering YouTube’s privacy policy exceptions for parody and satire.
Leslie Miller, YouTube’s vice president of government affairs and public policy, emphasized the platform’s commitment to safeguarding free speech, ensuring that videos with clear parody elements, even of world leaders, are likely to remain accessible. To participate in the program, individuals must provide a video of themselves along with a government ID. YouTube assures that this data will solely serve the likeness detection feature and can be deleted upon request.
Amjad Hanif, vice president of creator products at YouTube, noted that the number of removal requests under this policy remains minimal. This initiative underscores YouTube’s ongoing efforts to combat deepfakes and uphold privacy guidelines while preserving a space for diverse civic discourse.
Source: The Verge