In the era of rapid development of neural networks and generative artificial intelligence, the problem of unauthorized use of personal images is becoming increasingly acute. The video platform TikTok has begun testing a new tool designed to help creators find deepfakes created using their likeness. In the first stage, access to this feature will be granted to selected content creators in the United States.
How the verification system works
To gain access to the tool, test participants must undergo an identity verification procedure. This is carried out through the Jumio service, which requires the user to perform a document check and a real-time selfie. TikTok representative Zachary Kizer explained in an interview with The Verge that the Jumio service does not store user documents. Images are used solely for matching appearance and identifying possible cases of its unauthorized use.
Automatic search and reporting
After successful verification, the system automatically begins analyzing posts on the platform for the presence of AI-generated deepfakes. Creators have the opportunity to view the scanning results. If the system finds materials that violate the rules, users can file complaints about the corresponding videos or accounts.
Industry context
TikTok is not the only service tackling this problem. YouTube previously began implementing a similar tool, recently opening access to it for all adult users. The introduction of such technologies indicates that major technology companies are forced to adapt to new challenges created by generative AI and find ways to protect the digital portrait of their users.
