Sri Lanka
USA
Canada
India
UK
UAE
In a major step aimed at addressing the rise of AI-generated misinformation, YouTube is now extending its deepfake detection tool to political figures, civic leaders, and journalists, ahead of upcoming midterm elections. The move is intended to help public figures identify and manage content that misrepresents their likeness or voice.
Verified participants will be able to review videos flagged as using their likeness and request removal if they violate YouTube’s privacy policies. The proliferation of generative AI has made it increasingly easy to fabricate someone’s image or voice, raising concerns about misinformation and impersonation.
YouTube initially launched the tool in December 2024 for high-profile actors and athletes, later expanding it to top creators. Currently, around 4 million creators in the YouTube Partner Program are using it.
“We recognized the need to extend this technology beyond creators,” said Amjad Hanif, VP of Creator Products at YouTube. “Today, we’re starting a pilot program for journalists and government officials to help them protect their identities online. As we gather insights from election cycles and the way journalists use the tool, we plan to broaden access even further.”
Leslie Miller, VP of Government Affairs & Public Policy, added that the expansion aims to safeguard the integrity of public discourse, noting that the risk of AI impersonation is especially high for those in civic roles. While the company has not disclosed which leaders and journalists will be part of the pilot, it expects the program to grow rapidly and encourages policymakers and other stakeholders to get involved.
YouTube emphasized that the tool will not override principles of free expression. “Detection does not mean automatic takedown,” Miller said. “Parody, satire, and political critique will continue to be protected. A clear parody of a public figure, for example, is likely to remain online.”
Hanif also noted that among current celebrity and creator users, the number of removal requests has been relatively low. “While the tool highlights potential matches, most content is benign or even beneficial to the person’s overall presence,” he explained.
The platform has also released an explainer video outlining how the deepfake detection tool works and what eligible public figures can expect from its use.
