Spotify Doesn't Want to Get Rid of the AI Slop Clogging Up Your Playlists

True story: I like to listen to lo-fi ambient playlists while I work. It’s the ideal soundscape for getting stuff done, calming without being too distracting. But I’ve found over the past year or so that it’s become increasingly difficult to find a playlist or YouTube video of lo-fi music that isn’t full of AI-generated slop. I can hear it from a mile away (if the obvious uncanny Midjourney images attached to them weren’t a big enough red flag.) Now, it’s gotten to the point where, if I can’t locate a real human behind the channel, or they don’t actively note that their work is made without AI, I assume it’s AI. And that sucks.

But I’m sadly not alone on this journey of auditory confusion. Spotify has been at the forefront of polluting its users playlists and recommendations with vast swaths of AI slop. Many Spotify users have been vocal in their complaints, noting how nigh-on impossible it is to avoid this crap. No matter how carefully you curate your music, AI somehow finds its way in. And it seems that Spotify isn’t in much of a hurry to fix the problem. They’ve made a few concessions, mind you. In April, it launched, a test feature which shows, in a song’s credits, if an artist used AI. But that is a voluntary system that requires everyone involved to be honest, and AI slop merchants typically don’t want you to know that they’re tricking you.

Thousands of these AI slop tracks are uploaded to streaming platforms every day. They’re trying to outnumber the humans, to make it impossible to tell the difference. I imagine Spotify is not so secretly thrilled about the prospect of having fewer actual humans to pay out their $0.004 per stream royalties to artists. Unfortunately, this is becoming a new norm.

YouTube Music and Amazon Music have also avoided putting clear labels and filters on their platforms for users to avoid AI nonsense. By comparison, Deezer, a smaller platform, as noted by the BBC, “began both tagging albums that contain AI generated tracks produced by Suno, Udio and similar, and excluding the tracks from algorithmic recommendations or human-made playlists.” Apple Music said, external it was introducing “transparency tags” and were moving towards requiring disclosure from labels and distributors over whether or not they used AI. But, again, this seems to be voluntary, and slop houses don’t want to admit they’re selling you bilge from the plagiarism machine.

We’re already seeing a lot of the usual suspects try to claim that generative AI is just another tool they use to make work and that human ingenuity is still at the heart of their songs, but who believes that? These companies, like Suno, sell themselves as being able to “make” music with a few prompts and no heart, and they don’t want audiences to know the difference between real and fake. People have largely been in favour of transparency around the use of AI in “making” art, mostly because there’s no actual hunger for it and they don’t want to be sold a falsehood. If you couldn’t be bothered to make it, why should we be bothered to listen to read, to watch, and so on? I don’t come to art for your craven contempt, tech bros.