YouTube has announced that it will expand access to its AI-powered likeness detection tool to political and civic leaders as well as journalists, aiming to combat the spread of misleading or deceptive content generated using artificial intelligence. The move comes amid growing concerns about AI deepfakes and ahead of key election cycles in several countries.
Under the initiative, eligible politicians and journalists will be able to review videos identified as featuring their likeness and request removal if the content violates YouTube’s privacy policies. Advances in generative AI have made it increasingly easy to imitate a person’s face or voice, raising fears about misinformation online.
The platform first introduced the tool in December 2024 for well-known actors and athletes. It was later expanded to major creators, and YouTube says around four million creators in the YouTube Partner Program have now signed up to use the feature.
Amjad Hanif said the company had always planned to expand the technology beyond content creators. He noted that the initial rollout to journalists and government officials will begin with a pilot group to better understand how they use the tool to protect their identities online.
Meanwhile, Leslie Miller said the expansion is aimed at protecting the integrity of public discourse, noting that the risks of AI impersonation are particularly high for individuals involved in civic and political activities.
The company did not disclose which political leaders or journalists would be included in the pilot programme but indicated that participation is expected to expand quickly. Officials also said policymakers and other public figures are encouraged to contact the platform if they wish to learn more about the initiative.
YouTube emphasised that the tool will not automatically lead to the removal of videos. The company said it remains committed to protecting freedom of expression, meaning content such as parody, satire and political commentary will still be allowed on the platform if it clearly falls within those categories.
According to Hanif, although many celebrities and creators already using the tool detect numerous matches of their likeness online, the number of actual removal requests has remained relatively low, as much of the content is considered harmless or even beneficial to their public presence.
