Spotify investigates after activist group claims to have scraped 86 million songs from its library

Spotify has launched an investigation after an activist group claimed to have scraped tens of millions of music files and hundreds of millions of rows of metadata from the streaming giant. The group, known as Anna’s Archive, stated it has harvested 86 million audio files and 256 million rows of metadata in an effort to create a permanent "preservation archive" of humanity’s musical heritage.

The Stockholm-based company, which serves over 700 million users worldwide, confirmed it had identified and disabled the "nefarious user accounts" involved in the incident. Spotify characterised the activity as an "unlawful scraping" operation that utilised illicit tactics to go around digital rights management (DRM) and access the platform's audio inventory. Whilst Spotify hosts more than 100 million tracks, the company clarified that the leak does not represent or affect its entire library.

The incident has sparked alarm among copyright campaigners, who warn that the vast dataset is almost certain to be utilised by artificial intelligence companies for model data training. Ed Newton-Rex, a composer and advocate for artists' rights, noted that training on pirated material has become a "sadly common" practice within the AI industry. Observers suggest that such a dataset would allow developers to train music-generation models on modern, high-quality audio at an unprecedented scale.

Anna’s Archive, which has previously provided links to pirated books, argued in a blog post that its mission is to protect culture from "natural disasters, wars, and budget cuts." The group claimed its haul represents 99.6% of all music streamed by Spotify users and intends to share the files via torrents.

“Of course, Spotify doesn’t have all the music in the world, but it’s a great start,” the statement added.