TikTok launches global deepfake detection tool

TikTok has launched a global tool to help public figures detect AI-generated impersonations of themselves. The feature, called Likeness Detection, was announced on Friday 20 November and is being rolled out gradually to politicians, government accounts and accounts with large followings. It lets eligible users find AI-generated content that uses their image without consent and report it to TikTok.
The launch was announced at TikTok's safety and innovation event in London on Thursday, a day before the global rollout was confirmed. TikTok said the move is part of a wider effort to tackle AI-generated disinformation, with several European elections due in 2027. Business Today
How Likeness Detection works
Likeness Detection is opt-in. Eligible users who choose to sign up first have their identity verified through a trusted third-party provider. Once registered, they are notified when content or accounts on TikTok are detected as potentially using their likeness. They can then review the material and report it directly to TikTok's safety teams for assessment.
TikTok has stressed that identity documents used for verification will not be retained. The tool covers AI-generated images, videos and other content featuring a person's face, including material that may have been shared without their consent. It is being introduced in stages rather than switched on for every eligible account at once, and TikTok has not published a date by which all eligible users will have access. Switch Media
Why TikTok is acting now
The timing is tied to elections. TikTok said the tool is aimed at curbing AI-generated disinformation ahead of European elections next year, including contests in Italy, France and Greece. Francesca Scapolo, TikTok's government relations lead for European elections, said at the London event that protecting the integrity of elections was one of the platform's most important responsibilities, and that the tool would help politicians and government officials identify AI-generated material shared without their consent.
The launch also comes as the European Union begins introducing new AI transparency rules, including requirements for deepfakes and other AI-generated content to be labelled. TikTok has required creators to label realistic AI-generated images, audio and videos for several years, and the company said more than five billion AI-generated videos had been labelled on its platform.
A second transparency tool for everyone
Alongside Likeness Detection, TikTok introduced a separate transparency tool on 20 November that is open to anyone, including people without a TikTok account. It allows a user to check whether an image or video was created or edited using TikTok's AI tools. The check runs through TikTok's Transparency Centre and uses content credentials and an invisible watermark developed by TikTok to identify material made with its AI features.
TikTok is not alone in building this kind of protection. YouTube has introduced its own likeness-detection system to help creators identify AI-generated videos that may impersonate them. The wider question for all platforms is speed: detection tools and labels only help if harmful material is identified and removed quickly, particularly when it spreads across several platforms at once.
For now, TikTok has disclosed no figures for how accurately Likeness Detection identifies impersonations, and no full rollout timetable. What is clear is the direction of travel: public figures are being given a direct route to flag fakes of themselves, rather than relying on other users to report them first.
Sources: Business Today and Switch Media.