AI
TikTok rolls out AI deepfake detection for creators
The platform is testing a new tool that helps creators identify and flag unauthorized AI-generated videos using their likeness.
The Verge
TikTok is piloting an opt-in system designed to help creators spot and take action against AI-generated deepfakes of themselves. The company began testing the feature with select US creators, following similar moves by competitors like YouTube.
How the verification works#
Before accessing the detection tool, creators must confirm their identity through a third-party verification service called Jumio. This involves a real-time selfie and an ID verification step. According to TikTok's US spokesperson Zachary Kizer, the platform "does not retain ID documents," and any facial data collected is strictly used for comparing against potential deepfakes and flagging misuse.
Scanning for unauthorized AI content#
Once verified, TikTok's system automatically searches for AI-generated videos that may feature the creator's likeness. Creators can then review the results and flag any content or accounts they believe are using their image without permission.
Context with other platforms#
YouTube has been developing a comparable detection tool and recently expanded access to all adult users on its platform, suggesting this is becoming an industry-standard approach to combating deepfake abuse.
