YouTube is building tools to detect face and voice deepfakes

Estimated reading time: 2 minutes

YouTube is developing new tools to protect artists and creators from the unauthorized use of their likenesses. On Thursday, the company announced plans to roll out technology that can detect AI-generated content using a person’s face or singing voice, with pilot programs launching early next year.

The future face detection tool will enable people in different professions to ‘detect and manage’ the AI-created representations of their faces. To protect creators, actors, musicians, athletes and entertainers from deep fake use of their likenesses, YouTube is developing these tools but with no current plan for launch yet.

More like this

• OpenAI introduces tool for Identifying images generated by DALL-E 3
• ‘We take education seriously’, President Tinubu 
• How Tech software improves early Autism detection
• Tayo Aina, foremost Nigerian Youtuber explains 12 easy steps to establish a successful YouTube channel
• Kanayo O Kanayo addresses YouTube movie producers, tackles modern thespians ‘with no sense of purpose’

Also, a new capability called “synthetic-singing identification” will be included in Content ID which is YouTube’s IP protection system. This feature will assist partners in monitoring the checked singing voice AI versions.

“So as AI matures we think it should augment creativity, not replace it,” added YouTube’s VP and head of creator products, Amjad Hanif in a blog post. “We will remain loyal to our partners and will keep building defences to allow their ideas to be heard and make sure that the next evolutions are going to be even more inclusive. ”