YouTube announced on September 23 that it will expand its likeness detection tool to include voice detection alongside facial recognition. The update, set to roll out later this year, aims to improve match accuracy and provide broader voice protections for creators. The tool will also be available on the YouTube mobile app, allowing creators to enroll and receive alerts directly from their phones, according to medianama.com.
Currently, YouTube’s likeness detection scans newly uploaded videos for the faces of enrolled creators only. Creators who enroll must provide a government-issued ID and a selfie video, which YouTube stores for up to three years. The new voice detection feature will process audio data from creators’ content and may require a voice reference submission. Creators must consent to voice processing as part of enrollment, the company said in a blog post by Aparna Pappu, YouTube’s Vice President.
This enhancement comes amid growing concerns over AI-generated voice clones and likeness misuse, especially in India where courts have targeted AI voice cloning. By combining voice and facial detection, YouTube aims to strengthen protections against unauthorized use of creators’ identities. The move aligns with broader industry efforts to combat deepfakes and synthetic media, offering a more comprehensive safeguard for content creators’ personal likenesses.
YouTube’s updated likeness detection tool will be accessible via its mobile app, enabling creators to manage alerts and reports on the go. The company’s help page states that creators’ face and voice templates may be used to improve the tool, with data retention lasting up to three years from the last sign-in.