YouTube is working on an AI tool that will allow creators use artists’ voices

YouTube is developing an AI tool that will allow creators to use artists’ voices for their own content, Billboard reports. The three major labels — Universal Music Group, Warner Music Group, and Sony Music Entertainment — are all currently negotiating licensing deals with YouTube over voice rights for the beta version of the tool, a source told the magazine.

The tool was reportedly scheduled for a September beta launch. According to Billboard’s sources, though, it’s been difficult to find “top artists” who are willing to opt into the tool, with many worried their voices could be used for deceptive deepfakes that might harm their image or support values with which they disagree.

“Big Three” label executives, many of whom had previously voiced concern over the rapid rise of AI in music, have changed their tune of late. UMG lobbied Congress in July to regulate AI. But the company’s CEO, Lucian Grange, said in August that he believed AI could “amplify human imagination and enrich musical creativity in extraordinary new ways.” And Warner Music CEO Robert Kyncl spoke on AI’s inevitability last month. “You have to embrace the technology, because it’s not like you can put technology in a bottle,” he told Tech Crunch.

Key issues under negotiation, per Billboard, are how the tool is trained, artists’ ability to opt in or out, and the monetization model: Will artists be paid for the use of their music or for the content that’s made using the tool? As the publication notes, a public launch of the tool would require further negotiations.

The FADER has reached out to representatives of YouTube and all three major labels for comment.