The phenomenon of artificial intelligence (AI) imitating famous artists’ voices and styles in pop music has reached a tipping point. Fans can now easily create deepfake cover versions of songs using AI technology. Some consider this trend to be a fun diversion, while others view it as a potential issue. Regardless, record labels and tech platforms are seeking ways to monetize this development.
According to reports from the Financial Times, Google and Universal Music are currently negotiating a partnership to license music from Universal’s catalog for AI purposes. While no agreement or product launch has been confirmed yet, the aim is to enable fans to create AI-generated music based on the voices of Universal artists, with the copyright owners receiving compensation.
Warner Music is also engaged in similar discussions with Google, while Sony already has an executive dedicated to AI-related matters. This proposed system would allow artists to decide whether or not to participate. Grimes, for example, has been an early advocate of this concept by offering to share royalties with producers who utilize AI versions of her voice.
Label executives perceive the rise of AI in music as reminiscent of YouTube’s early days, where labels initially issued takedown orders against videos featuring their music. The situation changed over time, and now the music industry earns $20 billion annually from syncs in YouTube videos.
Earlier this year, a song called “Heart On My Sleeve,” which used AI to mimic the voices of Drake and the Weeknd, became a viral hit before Universal enforced its removal from streaming services. However, in a potential future with this new system, all stakeholders – including labels, tech companies, artists, and AI users – could profit from similar songs.
Warner Music’s CEO, Robert Kyncl, mentioned to investors that such a system could allow fans to pay tribute to their favorite artists through user-driven content, including cover versions and mash-ups. Google has already showcased generative AI software capable of creating music based on vague prompts, although it is not yet commercially available.