Spotify charts have been filled with AI songs. It is largely a consequence of what Spotify has encouraged
Songs like ‘I still breathe‘ either ‘I loved myself more‘, performed by Ruby Black, have topped Spotify’s algorithmic charts in Spain for weeks. Ruby Black does not exist: she is a singer generated with artificial intelligence distributed by a label called Silencio Capital, with more than one hundred thousand followers on Instagram and a new single every Thursday. Spotify is now announcing measures that try to quell this wave of synthetic artists, but everything indicates that this new situation is here to stay. And we have asked for it. Who’s that girl. In April 2026Ruby Black topped the first spot on Spotify’s list of The 50 Most Viral in Spain. Soul-type ballads, very soft, with vocal echoes of Rosalía and other fashion trends in pop in Spanish, lyrics of heartbreak and improvement, and covers and video clips generated by artificial intelligence. Google’s own AI, when consulted about the artist, lied when describing her as a “human, not AI, singer, known for ballads like ‘I Still Breathe’.” How are the machines? Its release of weekly singles is unequivocal: chain production applied to entertainment, thanks to a catalog that continually grows, minimal cost, zero emotional ties with real creators… Ruby Black is not the only one of her kind: the same viral top, as different media have commented, includes equally dubious artists such as Nyx Solaris. But Ruby Black is the one who has reached the highest thanks to the undoubted advantage of singing in Spanish. There is AI but… how much AI? According to what he said Deezer this monthreceive around 75,000 completely AI-generated tracks every day, 44% of all new content. In January 2025, just over a year ago, there were only 10,000. Spotify, for its part, removed more than 7.5 million tracks generated by AI in the twelve months prior to September 2025, many of them linked to mass generation tools. Its entire catalog is around 100 million songs. We don’t distinguish it. It is not only an ethical issue, but it is increasingly difficult to identify AI-generated music. A study commissioned by Deezer with 9,000 people in eight countries published in November 2025 concluded that 97% of the participants were not able to distinguish between songs generated by AI and human songs in a blind test. 80% of them believed that 100% artificial music should be clearly labeled. The demand exists but the platforms have not yet reacted. Step forward from Deezer. In fact, Deezer is the only one that, to date, has implemented its own detection system. In June 2025 began flagging albums that included 100% AI-generated tracks and to exclude them from algorithmic recommendations and editorial playlists. In January 2026 put that same technology on sale so that other platforms could take advantage of it. Deezer also found that 85% of streams of AI music are fraudulent, and that is why it excludes them from its distribution of royalties. It is the only one that has taken such an openly anti-AI stance. Apple Music launched in March 2026 Transparency Tagsa metadata system that allows labels and distributors to voluntarily declare whether they have used AI in vocals, songwriting, cover art or video. And Spotify works with DDEXthe music industry standards body, on a metadata system for song credits that indicates how AI has been used, also voluntary. Spotify’s latest move. To all this, the most followed platform has added the verification seal ‘Verified by Spotify‘ to ensure that there are humans behind each artist profile. Artists like Ruby Black are, precisely, proof that the formula limps: with a massive following on networks, a single every week, and number one on the top viral list in Spain, he has everything Spotify needs to award him the “real artist” label. A well-managed synthetic avatar can meet Spotify’s criteria (consistent activity, compliance with platform policies, signs of an artist’s real presence) and be a synthetic creation. The root of the problem. The truth is that all this fuss was not born with generative models. Liz Pelly, author of the book ‘Mood Machine: The Rise of Spotify and the Costs of the Perfect Playlist’, has been documenting for years how Spotify has systematically built a listening model based on lean-back listening: mood playlists, algorithmic recommendations that prioritize the most generic song in an artist’s catalog, the content that works best as a background sound. The author also revealed the existence of the internal program called Perfect Fit Content (PFC): since 2017, Spotify has filled its most popular playlists with “ghost artists” musicproduced in series to reduce the cost of royalties. Twenty composers were behind the work of more than five hundred “artists”and its tracks were listened to millions of times. Playlists such as Deep Focus, Cocktail Jazz or Ambient Relaxation were almost entirely composed by Perfect Fit Content. AI has not broken into a healthy ecosystem: platforms have been favoring anonymous, interchangeable and depersonalized content for years. And awarded by the algorithm. The Instagram-core. The phenomenon has an exact reflection on Instagram. In the same way that there is a Instagram-core (that homogeneous aesthetic of Reels with fast transitions, viral music, warm light and motivational text), there is almost the Spotify-core that Ruby Black represents. That is, designed to exceed the thirty-second hearing threshold (quick emotional hook, recognizable or directly cloned voice, lyrics that impact at first) that Spotify counts as a listen. Or put another way: Spotify can delete 75 million tracks and announce anti-spam filters. But as long as it continues to reward with its algorithm the most comfortable, ephemeral and generic song to captivate a listener who listens without paying attention, Ruby Black is just one more profile (a millionaire, of course) in a problem that has been brewing for years. In Xataka | Dear Spotify, it’s about time you had a button that allows you to filter AI-generated music
