Pirate activist group Anna's Archive says it scraped 86M music files and 256M rows of track metadata from Spotify, and releases them in ~300TB of torrent files
The breach allegedly includes 86 million audio files.
BillboardMitchell Peters
Context & Ripple Effects
Anna's Archive's claimed torrent release is the opening move in a dispute that quickly became a rights-enforcement case: Spotify and music labels later sued over the planned distribution, obtaining a temporary order against some domains in related coverage.
The claimed archive did not remain merely a dataset announcement. Related coverage says the group later began releasing millions of tracks from the scrape despite a preliminary injunction, and the case ultimately produced a $322.2M judgment that Billboard characterized as difficult to enforce against an anonymous operator.
First-order effects
The alleged release makes a large collection of music files and accompanying track metadata available outside Spotify's licensed access model, directly affecting Spotify, labels, and rightsholders whose works are included.
Spotify and labels are pushed toward rapid legal and domain-level responses; the later lawsuit and temporary domain takedown order show that this became an immediate enforcement priority.
Second-order effects
Rightsholders and streaming services face a broader operational burden: they must pair litigation with measures intended to limit mass extraction and redistribution, while infringement can persist across torrent distribution and replacement domains.
The inclusion of extensive metadata raises the stakes beyond audio copying, because structured catalog information can make a redistributed collection easier to search, organize, and reuse.
Third-order effects
If large-scale scraping and decentralized release remain repeatable, music licensing enforcement will increasingly hinge on limiting access and distribution pathways before files spread, rather than relying on damages after release.
The case also illustrates a persistent enforcement gap: large judgments can establish liability without necessarily restoring control when the accused operator is anonymous or outside easy reach.
The trend: This is part of a shift in which streaming platforms and music rightsholders must treat catalog-scale data extraction as both a copyright-enforcement and distribution-control problem.
The metadata is super interesting... ~37% of songs ≈ 99.6% of listens, so the remaining ~63% of songs ≈ 0.4% of listens. (They note Spotify doesn't publish stream counts for songs with <1000 streams, forcing estimation/error bars in the distribution part w “≥70% of songs.")
The top 3 songs have a higher listen count than the bottom 20-100 million songs COMBINED 70% of the songs on Spotify are ones that no one ever listens to. 37% of all songs represent 99% of the listens https://annas-archive.li/... [image]
Wow! “We backed up Spotify (metadata and music files). It's distributed in bulk torrents (~300TB), grouped by popularity. This release is the largest publicly available music metadata database with 256 million tracks and 186 million unique ISRCs” Link: https://annas-archive.li/..…
Fun stat: all of Spotify ≈ 300TB. Here's another fun one at Dabba we move 3× that that every day from normal internet usage. ≈ 870 TB in the last 24 hrs Scale hits differently with familiar benchmarks.
“We backed up Spotify (metadata and music files). It's distributed in bulk torrents (~300TB), grouped by popularity. This release includes the largest publicly available music metadata database with 256 million tracks and 186 million unique ISRCs.” https://annas-archive.li/...
ALL the songs on Spotify were scraped, and released for free as a 300TB torrent. Artists won't get paid, obviously, so this is hardly true ‘activism’. Meanwhile, OpenAI lawyers might well be muttering “sounds legit”... www.billboard.com/business/str...
This doesn't help the artists, so this pirate activist group sucks, whatever their intentions may be. — I'm good with boycotting Spotify. (I don't use a streaming service myself. I buy CDs like a good Luddite.) [embedded post]