Third-party trackers hunt down Spotify’s mystery AI music stars

Slime Dot was supposed to be Las Vegas's next big R&B act. She had over 100,000 monthly listeners on Spotify and even posted a photo with Drake. When rumors swirled in May that she was purely artificial, GQ ran an interview credulously helping her debunk the AI claims.

Then came the awkward editor's note. GQ had to admit she was an AI avatar after all, sheepishly adding that the talent behind the music remains undeniable.

With Spotify dragging its feet on labeling generative music, the community is taking matters into its own hands. Two independent tools, SoullessMusic and SlopTracker, have emerged to hunt down synthetic audio across the platform.

The bounty hunters tracking audio slop

Frustrated by seeing human musicians get ripped off and AI clones popping up across streaming services, developer Graeme Fulton launched SoullessMusic. The site uses open-source detection models like SONICS, vocoder fakeprint detectors, and metadata scrapers to catalog what it calls soulless melody.

SlopTracker takes a direct approach by letting anyone paste a Spotify link for immediate analysis. It found that Slime Dot's track Fully had a 95 percent probability of coming straight out of Suno.

The real issue is cash. SoullessMusic estimates that the handful of AI artists it tracks siphon roughly $5.7 million in streaming payouts every year. The top AI persona, mikeeysmind, pulls in an estimated $1.5 million annually by himself.

SlopTracker calculates that AI tracks landing on Spotify's official curated playlists bleed about $0.1188 every single second away from human musicians. It gets worse: scammers have uploaded AI-generated songs under the names of dead musicians to harvest streams, while anonymous accounts like Qajar Jazz churn out hours of synthetic Iranian jazz each week.

Spotify claims it is using a layered approach with Verified artist tags and voluntary AI credits, pointing to tens of thousands of daily disclosures. But across standard playlists, synthetic tracks routinely slip through unflagged.

Why it matters

This is not a technical limitation; it is an incentive failure. Generative engines like Suno and Udio already bake inaudible digital watermarks into their audio outputs, meaning Spotify could easily implement automated labeling if it truly wanted to. Competitor Deezer revealed in April that a staggering 44 percent of all new uploads are already AI-generated, proving just how massive the flood has become.

For listeners, you are paying subscription fees for human artistry while algorithms feed you cheap synthetic wallpaper. For working musicians who already earn fractions of a cent per stream, the streaming royalty pool is zero-sum. Every dollar swallowed by a fake persona pumping out thousands of prompt-based tracks is a dollar stolen from real artists paying rent and studio time.

We think Spotify's reluctance looks like a quiet hedge. AI music does not come with major label bargaining power or fussy royalties, which could theoretically lower platform payout obligations over time. But letting independent hobbyists police basic authenticity is an awful look, and trading artistic trust for short-term catalog bloat is a dangerous game.