‘We Backed Up Spotify’: Shadow Library Unleashes 300TB of Music Data in Torrent Shock

Anna's Archive, the world's largest shadow library—and a platform increasingly funded by AI developers—sent shockwaves across the internet this week by announcing it had "backed up Spotify." The group claims it has begun distributing 300 terabytes of music metadata and audio files via bulk torrents, a move that challenges copyright norms on an unprecedented scale.

The Scale of the "Preservation Archive"

According to Anna's Archive, the massive data grab represents "the largest publicly available music metadata database," containing 256 million tracks. The archive boasted that this dataset covers more than 99 percent of all listens on Spotify.

The release also includes 86 million actual music files, which supposedly represent about 37 percent of the songs available on Spotify as of July 2025. The group framed the project as "the world’s first 'preservation archive' for music which is fully open."

A Curated Cache for the AI Era?

This wasn't an indiscriminate data dump. Anna's Archive stated that the scraped files were prioritized by popularity. It also took steps to weed out songs that are never streamed or are of poor quality, explicitly mentioning the removal of AI-generated songs. This curation suggests an intent to create a high-quality, structured dataset.

The source's mention that the archive is "increasingly funded by AI developers" is critical. High-quality data is the lifeblood for training AI models, and tech companies face immense legal and financial hurdles in acquiring it legitimately. This release could be seen as a direct challenge to the copyright frameworks that AI companies are struggling to navigate.