A live performance from Ringo Starr, one of the many artists whose voices are seemingly being offered as AI-powered soundalike options on Jammable. Photo Credit: slgckgc
The British Phonographic Industry (BPI) is threatening legal action against the owner of an AI soundalike-voice platform called Jammable.
London’s BPI today confirmed the development on social media, pointing readers to a (paywall-blocked) breakdown of the situation from The Times. Jammable (formerly Voicify AI, which is now Jammable’s parent) bills itself as “the #1 platform for AI music.”
And according to its website, the service enables users to replicate uploaded voices, automatically remove reverb from recordings, generate text-to-speech vocals, and make existing projects’ vocals sound as if they’d been recorded in different voices, among other things.
Once again per the Jammable website, these custom-audio creations can be made to sound like recordings from various actors, politicians, celebrities, and cartoon characters.
Predictably, given the BPI’s threat of litigation, London-based Jammable also allows customers to generate audio with soundalike artist voices, including but not limited to Eddie Vedder, Phil Collins, Ringo Starr, Daddy Yankee, Ice Cube, and Chris Martin, the website shows.
According to the mentioned Times report, it’s the latter offerings, presumably trained on the appropriate professionals’ real-life voices, about which the BPI has expressed infringement-related qualms. “Formal legal action” could arrive unless Jammable admits to the claims and curbs the alleged infringement, per the outlet.
DMN reached out to Jammable for comment but didn’t receive a response in time for publishing. On the opposite side of the dispute, longtime BPI general counsel Kiaron Whitehead relayed in a statement: “Music is precious to us all, and the human artistry that creates it must be valued, protected and rewarded.”
Bigger picture, the Jammable-BPI dispute represents one component of a broader industry crackdown on unauthorized AI projects, which made all manner of headlines last year.
To this point in 2024, far-reaching layoffs, the Universal Music-TikTok dispute, legal actions over artificial intelligence companies’ training data, and more have in many ways pushed audio deepfakes out of the media spotlight.
But it was only in December that Universal Music indicated AI fraudsters were selling fake pre-release tracks for as much as $30,000 a pop – with Sony Music disclosing that it’d “sent close to 10,000 takedowns” over AI deepfakes.
Against this backdrop, 2024’s start brought with it the introduction of the No AI Fraud Act in Congress, proposing fines for unapproved soundalike releases, a federal right of publicity, and more. And earlier in March, the EU passed the AI Act, establishing (among an array of other things) mandatory training disclosures for AI systems. However, the AI Act won’t become “fully applicable” until about two years from now.
Worth mentioning in conclusion is that Sony Music, which opted not to participate in the first iteration of the Universal Music-spearheaded AI partnership with YouTube Shorts, previously said it intended to release an artificial intelligence project that’s “very much an experiment and a first of its kind.”
Shortly thereafter, the major label unveiled a partnership with Vermillio, which says it’s “the first and only generative platform built to empower artists.”
