Spotify recently purged tens of thousands of AI-generated songs from its vast digital music library. The problem doesn’t appear to have been the songs themselves—which were generated using a proprietary AI software. Instead, Spotify was concerned about the listeners—who, as it turns out, were also generated by software.
The Financial Times reports that Spotify temporarily cut off new uploads from the AI music startup Boomy last week and took down thousands of its songs due to suspected automated streams of Boomy’s music.
Boomy has been a big player in the AI music industry for a couple years now. Launched in 2021, the company offers users the ability to auto-generate an assortment of tracks based on particular “vibes.” Say you want to spin up a hip hop track with Reggae undertones? Just input those requests into Boomy and ta-da, you have yourself an “original track.” Users can then upload Boomy tracks to streaming platforms and earn royalties from them.
Predictably, it would appear that some people have been taking advantage of this automated music model to generate fraudulent streaming traffic that can then be illegitimately monetized. It’s an infinite listening loop that needs no human ears. FT reports that Spotify took action against Boomy due to suspected bot activity that was spiking listener levels for some of its tracks. The activity was flagged by the Universal Music Group, a giant label representing some of the biggest stars in the industry, that recently called for streaming services to crack down on AI content. Spotify took down approximately seven percent of the Boomy tracks—amounting to “tens of thousands” of songs.
When reached for comment by Gizmodo, Spotify confirmed that it had recently taken action against Boomy due to the detection of “artificial streaming.” A Spotify spokesperson provided the following statement on the issue of fake streamers:
“Artificial streaming is a longstanding, industry-wide issue that Spotify is working to stamp out across our service. When we identify or are alerted to potential cases of stream manipulation, we mitigate their impact by taking action that may include the removal of streaming numbers and the withholding of royalties. This allows us to protect royalty payouts for honest, hardworking artists.”
While Boomy was kicked off Spotify last week, it appears the startup has since been welcomed back onto the platform. In a statement shared on its Discord channel this week, the company said it was “pleased to share that curated delivery to Spotify of new releases by Boomy artists has been re-enabled.” The company has said that it is “categorically against any type of manipulation or artificial streaming.”
If automated streams aren’t a particularly new problem for streaming platforms, the advent of AI music certainly seems to make this problem all the more ridiculous. Fake audiences listening to fake music? You’d be hard pressed to find a weirder, more dystopian form of fraud than that.
The incident is also a sign that the “generative AI” revolution is causing a lot of weird headaches for the music industry, though it depends on who you ask. If you ask ethereal synth-pop elf Grimes, who recently launched a platform that lets you use AI to simulate her voice to create new music, she’d tell you that AI music is just a beautiful evolution of the art form. But if you can ask someone who’s less thrilled with the tech, like Drake, he’d probably tell you he’s certifiably pissed that an online creator recently made an AI song featuring his voice without his consent. And, of course, if you can ask the Universal Music Group, they’d tell you that AI music as “fraud” and and that it should be banned from streaming platforms.
The controversies surrounding AI music are part of a broader pattern of backlash and concern dogging the new innovations in automation. AI acolytes have promised that their technologies will change the world for the better but, so far, their products seem to be mostly just filling the internet with synthetic crap.