YouTube tackles deepfakes with new detection tool, but privacy concerns mount
YouTube is rolling out a new tool designed to help creators protect themselves against deepfakes that manipulate their faces. The lookalike detection tool scans uploaded videos for artificially altered faces and lets creators request takedowns. But the way YouTube plans to verify users’ identities is already drawing sharp criticism.
To use the new feature, creators must submit a government ID and a biometric video of their face. That raises questions about how safely this highly sensitive data will be handled, US news channel CNBC reported.
YouTube stressed that the biometric information will only be used to confirm identities and bolster security. Even so, experts pointed out that the data would still fall under Google’s broader privacy policy. In practice, that means publicly available biometric information could be fed into AI training models.
Security firms such as Vermillio and Loti have advised creators to proceed with caution. They warned that giving up control over biometric data carries significant risks, especially now that a face is increasingly treated as a valuable piece of digital property. Once that control is lost, they argue, it is nearly impossible to get it back.
Critics are particularly worried that Google could repurpose users’ facial data for its own AI development, without explicit consent or compensation.
YouTube maintains that the tool is specifically intended to shield creators from misuse of their likeness. The detection feature is expected to be rolled out by the end of January to millions of creators in the YouTube Partner Program. According to the platform, the tool should help creators safeguard their reputation and maintain audience trust at a time when deepfakes are surging.
Still, concerns persist. Creators fear there is no way to profit from unauthorised commercial use of their face, and some are questioning how reliable the detection system really is. False positives could, after all, lead to videos being taken down unfairly.
YouTube said early tests have been encouraging and that most creators, even when their content was flagged as suspicious, ultimately chose not to have the videos removed.
Business AM
Subscribers 0
Fans 0
Followers 0
Followers