Music streaming services like Spotify and YouTube Music have become the primary way people listen to music, which can be a lot more convenient than buying individual albums. However, this also makes it easier for AI-created tracks to worm their way into your playlists. Most streamers don’t go out of their way to label AI music, but Deezer has worked to develop technology to identify that content. In a recent update, the company says AI music is approaching half of all new uploads, and most of the supposed listeners of those streams are AI themselves.
AI-generated music has taken off in the last few years, but it doesn’t get as much attention as other parts of the AI ecosystem. That’s due, in part, to the fact that AI music can fly under the radar. With the right context and prompting, an AI track can sound like generic, over-produced music created by humans. According to Deezer, its users have a hard time differentiating AI tunes from the real deal. Listeners taking a Deezer survey listened to three songs, two of which were AI. A whopping 97 percent were unable to tell the difference between the AI songs and the one made by a human, the company reports.
Deezer says it has developed technology to detect AI uploads, and it’s one of the few streamers to explicitly label such content. As generative audio models have proliferated, the rate of AI uploads to Deezer has reached a staggering 44 percent—that’s 75,000 new AI tracks on Deezer every single day. Deezer licenses this technology to third parties, which it claims has a false positive rate of less than 0.01 percent.
Still, listeners shouldn’t encounter AI music organically on Deezer because the site won’t include AI-flagged tracks in suggestions or editorial playlists. As a result, AI music streams account for a small share of Deezer usage, hovering around 1–3 percent. The company says the primary purpose of uploading all this AI music is fraudulent. Deezer only pays for streams when a person listens to them, so it’s demonetizing about 85 percent of AI music streams.
“Thanks to our technology and the proactive measures we put in place more than a year ago, we have shown that it’s possible to reduce AI-related fraud and payment dilution in streaming to a minimum,” said Deezer CEO Alexis Lanternier.
The AI goes to 11
The growth of AI music is likely to continue accelerating with the rest of the AI industry. Models like Google’s Lyria 3 have become cheaper and more widely available. Google actually lets Gemini users generate full-length songs now, up from 30-second snippet tracks just a few months ago. Suno and Udio also promote their ability to create broadcast-ready tracks in seconds. However, these mainstream options embed watermarks, like Google’s SynthID, to flag the songs as AI.
The problem is how easy it is becoming to strip those watermarks out of the audio and generate music using custom models that don’t have them in the first place. As AI inference becomes cheaper, so, too, does the creation of musical AI slop.
