Figaro will help streaming services deal with AI

UK-based audio search and detection startup Figaro wants to help streaming services to deal with the nuanced challenges of AI-generated music.

The company is responding to the recent industry debates about this topic by extending its audio-search technology into AI content moderation.

CEO Lydia Gregory told Music Ally that its focus is on helping DSPs to tackle three separate (but related) issues with content: quantity, quality and copyright infringement.

The first concerns the sheer amount of tracks being uploaded to streaming services every day, which has grown from an estimated 20k a day in 2018 to around 120k now. With the prospect of generative AI technologies swelling that number further, services face the challenge of making all that music searchable.

When it is easier than ever to both make and release music, DSPs also face the challenge of trying to ensure that they’re only surfacing good-quality content to their listeners.

The quality of music is subjective, of course, and Gregory stressed that Figaro will not be making itself the judge of ‘good’ and ‘bad’ music. Instead, it will work with each of its streaming-service clients to identify which music does not fit the expectations of their listeners.

The third prong of Figaro's expansion is helping streaming services to identify music that may infringe artists’ rights: for example as seen in the recent rash of voice-cloned tracks released without the permission or involvement of the artists.

"I don't think it makes sense, or is practical, to achieve a blanket ‘is it generative AI?’ detection system, and so we’re considering solutions for each one of these challenges," said Gregory, who before she co-founded Figaro (which was originally called FeedForward but has since rebranded) worked as head of growth for one of the first AI music startups, Jukedeck.

"If automated detection is used, there's then a question about what a business wants to do with that information. Take down? Pay royalties? Don't surface in search? Or something else? Working that out will require collaboration between rightsholders, artists, platforms, and technology experts."

Figaro is also keen to highlight the fact that generating audio is just one use for machine-learning technology, and that (as with all tech) it is a tool, so working out how to use it for the benefit of music companies and artists is an important task for the industry and startups/developers to work together on.

News of Figaro's expansion comes shortly after streaming service Deezer announced that it was planning to build more technology to identify AI-generated music, and ultimately to "develop a remuneration model that distinguishes between different types of music creation" based on that.

Music Ally's next Learn Live webinar will help you build the strategies for artists to thrive in new international markets!

Music Ally's Head of Insight More by Stuart Dredge