Spotify won't label AI music, so fans built SoullessMusic and SlopTracker to do it

Spotify does not tell you which songs on its platform were made by AI. Two volunteer-run websites now do it instead, and they have built the databases Spotify has spent two years declining to publish.

SoullessMusic.com and SlopTracker.org both scan Spotify, score tracks for AI generation, and publish the results with names, listener counts and earnings estimates attached. 404 Media reported on the pair this week.

What SoullessMusic and SlopTracker actually track

SoullessMusic runs like a directory. Each act gets a confidence score, a monthly listener count, an estimated monthly payout and a verification date. The site is open source and also collects submitted stories from musicians and listeners who say they were fooled.

SlopTracker frames the same data as a loss. It counts 42 AI artists, roughly $2.13 million in estimated earnings, $244,059 in expected monthly income and 25,913 premium subscriptions absorbed per month. It also runs a scanner where anyone can paste a Spotify track link and get a verdict.

The gap between 232 and 42 is the first thing to notice. Two sites looking at the same platform disagree by a factor of five on how many AI artists exist, because each picked its own threshold and its own method. No standard exists for either of them to follow.

The Slime Dot case that pushed GQ into a correction

Slime Dot is the example that made this concrete. The R&B act has more than 100,000 monthly Spotify listeners and spent months denying it was AI, including in a GQ interview in May 2026 published under the framing that it debunked the rumours.

“

The truth doesn't matter these days. People are always gonna try to explain something they don't fully understand and believe what they want.

GQ later added a note to the piece conceding Slime Dot was an AI avatar, while adding the talent behind the music “remains undeniable.” Both trackers flag the account.

The magazine had no tool to check. Neither did the readers. Two hobbyist databases did.

Why the same track gets labeled on YouTube and not on Spotify

404 Media also surfaced Qajar Jazz, a channel that appeared in December 2025 and now publishes hours of Iranian jazz every week to 29,579 monthly Spotify listeners. Both trackers detect it as AI-generated.

On YouTube, the same catalogue carries an AI disclosure in the fine print. On Spotify it carries nothing. Comments on the YouTube videos suggest many listeners still believe real musicians made it.

That split is the actual story. The disclosure exists, it travels with the upload on one platform, and it evaporates on the other. Spotify’s announced policy work on AI slop and fake artists has not produced a consumer-facing label, and the RIAA and IFPI tagging proposal remains voluntary with no streaming service committed to displaying it.

Where the volunteer databases break

Look at who SoullessMusic lists. Xania Monet appears at 524,000 monthly listeners with a 99 confidence score. Breaking Rust appears at 974,000.

Neither is hiding. Xania Monet is a publicly credited AI project with a human songwriter behind the lyrics, signed in a deal reported across the trade press. The database sorts it into the same bucket as an account built to pass as human.