Technology news and discussion on TalkFuse.
TikTok is testing an opt-in ‘Likeness Detection’ tool that lets US creators find and report AI deepfakes of themselves
TikTok is testing a new opt-in feature called "Likeness Detection", which allows creators to find and report AI-generated deepfake videos using their face without permission. The tool, currently limited to a small group of US creators, marks TikTok's latest effort to tackle AI impersonation as deepfake technology becomes more widespread, following similar moves by rivals such as YouTube.
To use the feature, creators verify their identity via third-party service Jumio, using a real-time selfie scan and government ID check; TikTok's systems then scan the platform for matching AI-generated content, which creators can review and report. TikTok says it does not retain ID documents and only uses facial data for likeness matching. The launch follows TikTok's announcement that it has labelled over 3 billion AI-generated videos, removed more than 86 million fake accounts in early 2026, and joined the C2PA steering committee on content authenticity standards. YouTube introduced a comparable likeness-detection tool in October 2025, later extending it to celebrities, talent agencies and all adult users by May 2026.
In short
- TikTok trials "Likeness Detection" for US creators to spot AI deepfakes of themselves.
- Requires ID verification via Jumio; face data used only for matching, not stored.
- Follows TikTok's 3bn AI-labelled videos and mirrors YouTube's existing tool.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
This tool represents necessary progress against a genuine and escalating harm. Deepfakes can damage creators' reputations, enable harassment, and spread disinformation; offering creators the means to detect and remove such content addresses a real problem. The opt-in design respects autonomy, third-party verification through Jumio provides security safeguards, and alignment with industry standards like the C2PA demonstrates commitment to content authenticity. Creators gain concrete ability to defend their own likeness.
The case against
Whilst protective in intent, this approach sacrifices privacy for limited benefit. Creators must surrender biometric data and government ID to participate, creating new privacy risks that may outweigh gains, particularly given existing concerns about technology firms' data practices. The tool remains reactive rather than preventative, detecting only existing deepfakes rather than stopping creation, and remains accessible only to verified creators, leaving ordinary users—who may be most vulnerable to impersonation—largely unprotected.
Explained simply
AI technology can now create fake videos of real people by using their face, often without permission. This has become a growing problem online as the technology has become easier to use and more convincing.
TikTok is testing a new tool that helps creators find and report when AI-generated videos using their face appear on the platform. To use it, creators verify their identity with a selfie and ID check, after which TikTok scans for matching fake videos.
This effort is part of a wider industry response, as other social media platforms like YouTube have already introduced similar tools. TikTok says it has taken other steps to address the problem, including labelling billions of AI videos and removing millions of fake accounts.
AI Americas Technology Trending Weird & Viral World