I opened my Spotify app last Monday morning, expecting the usual dopamine hit of my Discover Weekly, but something felt off.
I hoped to find a new band. Instead, I got a track that sounds like elevator music. I skipped it, but the next track sounded similar.
My Discover Weekly is terrible now, and I know I’m not the only one. Paying subscribers are furious. Hop onto any forum, and you see the same complaint across the internet.
Discover Weekly is full of AI slop. People are using generative AI to mass-produce thousands of tracks and siphon money from the royalty pool.
If you fail to spot this garbage, you train Spotify to think you love it.
Is AI slop invading your YouTube Music recommendations? You’re not alone
First Spotify, now YouTube Music
Why AI artists are cashing in on Spotify streams
To understand the flood, follow the money.
Spotify operates on a pro rata royalty model. The company pools the subscription and ad revenue, keeps 30%, and distributes the remaining 70% to rights holders.
Human musicians spend months writing a hit song to earn a living. Scammers take a different route. To make real money at fractions of a cent, they need catastrophic volume, so they go after distracted listeners instead of passionate music fans.
They focus on functional genres such as sleep sounds, study beats, white noise, and background jazz. People put these playlists on while they work, study, or sleep, completely ignoring the audio.
The stream count ticks up hour after hour with no human stopping to judge the music.
How to spot a fake artist before you press play
Your Discover Weekly needs defense. Here’s how you spot a fake artist before they infect your feed.
Blank bios are your first warning sign
Real artists are desperate for your attention. They operate in a brutal attention economy. They have active Instagrams, TikToks, and detailed tour dates. The AI slop artists are digital ghosts.
Click the artist profile. The immediate red flag is an empty bio, or if text exists, it reads like a ChatGPT hallucination. You will see hollow phrases like “connecting the world through music” or “an endless quest for sonic perfection.”
AI album art is giving away fake tracks everywhere
Stream farmers automate the audio and the packaging. They use generative image models for their album artwork. Look for the telling signs of typical AI art.
Watch for visual artifacts and mangled text with gibberish letters. Check the hands for six or seven fingers. Look for objects melting into the background or instruments that make zero anatomical sense.
Independent musicians despise this aesthetic. Releasing music with tacky AI art is a massive red flag for artistic integrity.
Too many albums too fast? It’s probably AI
Humans require time to write, record, mix, and master audio. AI needs seconds. Check the artist’s discography. High-volume, rapid-fire releases signal automated generation.
Watch out for artists who put out five albums this year but weren’t around in 2023. People caught the AI entity Aventhis dropping three albums totaling 57 tracks in four months.
Another fake profile, Sienna Rose, uploaded over 45 songs in eight weeks. Taylor Swift works hard, but even she can’t drop five albums in a year (except maybe after a breakup).
Scammers exploit the 30-second stream rule
A stream counts as money only if it plays continuously for 30 seconds. Scammers optimize for this metric. They keep their tracks short to maximize the number of paid plays they can squeeze into a single hour of background listening.
Check the track lengths on the album. The giveaway is a playlist where every song sits right in that 31 to 40 second sweet spot. Bad actors cut out intros and musical flow, leaving a blunt audio made to hit the monetization floor.
Use metadata to spot generative audio
Spotify offers a built-in lie detector. Right-click the song on your desktop app — or tap the three dots on mobile — and select Show Credits.
Real music is collaborative. A genuine track lists distinct human names across the Written by, Produced by, and Performed by categories.
AI slop exposes itself here. The metadata often lists the artist’s name across all three fields, or attributes the entire creation to a shadowy, un-Googleable corporate alias.
Spotify relies on labels and distributors to supply this metadata. When stream farmers bulk-upload 10,000 tracks, they rarely bother inventing realistic human collaborators for the credits.
Fix your Discover Weekly in two easy steps
If your Discover Weekly already sounds like a robot, you need to execute a hard reset. Block the artist.
Open the fake artist’s page, tap the three dots, and select Don’t play this artist. This scrubs their audio from your personalized suggestions and playlists.
More importantly, it sends a negative signal to Spotify’s algorithm.
Next, turn on a Private Session before you play your functional tracks. That way, your eight hours of sleep sounds don’t mess with your taste profile.
