TikTok is implementing a system to search for AI-generated copies of bloggers' faces: a new frontier in protecting copyrighted content.
The popular short-video platform TikTok is launching a test of an innovative tool called Likeness Detection, designed to identify and block deepfakes created using content creators' faces. Currently, the feature is available to a limited group of creators in the United States.
The essence of the technology is simple yet effective: the system automatically scans uploaded content for matches with a specific blogger's face, after which the creator receives a notification and the ability to file a complaint about a suspicious deepfake or fake account. This is a step in the right direction — the problem of unauthorized image use in the era of generative neural networks is becoming critical.
How It Works
To activate the feature, a creator must undergo identity verification through a third-party service called Jumio. The procedure involves two steps: a real-time selfie and scanning an identification document. It is important to note that TikTok, according to the company, does not store copies of documents, and biometric data is used solely to find matches in content.
In my opinion, this approach is a reasonable compromise between the need to protect copyright and privacy requirements. However, the key question remains open: how accurate is the system in an environment where deepfakes are becoming increasingly realistic?
My expert assessment: The introduction of Likeness Detection is not just a technical innovation but a signal to the entire industry. Social platforms are beginning to realize that without active measures to protect creators' digital identities, reputational and legal risks will only grow. However, it is too early to rely solely on automated systems — a combination of algorithms and manual moderation will be needed, especially in gray areas where AI copies are used for parody or educational purposes.