YouTube Launches AI-similar Detection Tool To Fight Deepfake Content

JAKARTA YouTube officially launched a likeness detection technology on its platform. This new tool can be accessed by eligible creators in the YouTube Partner Program.

With this technology, creators can request the removal of content made by Artificial Intelligence (AI) that illegally utilizes their faces or voices. This is a type of content commonly referred to as deepfake.

This is the first batch launch after Apple tested its detection tool for a long time. For a number of eligible initial creators, they will receive eligibility emails.

This similarity detection technology works by identifying and managing content created by AI. Specifically, the detected content is content that displays faces and sounds that are similar to creators.

This feature is designed as a form of defense of the misuse of a person's photos and voices. Usually, the faces and voices of well-known creators are often used to promote a product or service without consent and disseminate disinformation.

To start using this tool, creators need to open the 'Likeness' tab on their channel and agree to data processing. This process is followed by scanning the QR code using a smartphone for identity verification.

This identity verification requires uploading official identity photos and selfie videos. Without verifying, this detection tool will not be usable. Once access is provided, the creator can manage the detected content.

Creators can view all videos detected by the system and then send deletion requests. This option can be exercised based on YouTube privacy guidelines or apply for copyright removal.

YouTube has been testing this technology since the start of this year and previously partnered with Creative Artists Agency (CAA) in identifying AI content. The platform also expressed its support for the NO FAKES Law in the United States which focuses on malicious AI replicas.