Deezer is adding an opt-in toggle that lets listeners filter AI-generated music out of their recommendations and feeds, with the rollout promised before the end of the year — the first major streaming platform to offer listener-side AI content control.
Key takeaways
- Deezer will roll out an AI-generated music filter toggle to users by the end of 2024.
- The feature is opt-in, meaning AI tracks remain available by default; listeners must actively enable the filter.
- No other major streaming platform — Spotify, Apple Music, Amazon Music — has announced a comparable listener-facing AI filter.
- The move comes amid rising legal pressure on distributors, including a UMG lawsuit against DistroKid over alleged AI-generated content flooding its catalog.
- For AI musicians and producers, this creates a new discoverability variable: whether your track is flagged as AI-generated will now directly affect who hears it.
What Deezer's filter actually does — and doesn't do
According to TechCrunch, the toggle is opt-in: listeners who want to exclude AI-generated tracks must switch it on themselves. AI music stays in the catalog and remains fully playable — this isn't a platform-wide ban. What it does is remove flagged tracks from algorithmic recommendations and curated feeds for users who choose that option.
The practical consequence for AI music creators is real and immediate. If Deezer's detection or metadata system tags your release as AI-generated, enabling the filter across even a fraction of its user base shrinks your algorithmic reach on the platform. How Deezer identifies AI-generated content — self-reported metadata from distributors, automated detection, or a combination — hasn't been fully detailed, which matters enormously for producers who blend AI-assisted composition with human performance.
The metadata problem nobody is solving yet
The filter's effectiveness hinges entirely on accurate labeling. Right now, music distributors like DistroKid and TuneCore don't have a standardized, mandatory field for AI-generated content — some platforms have added voluntary disclosure options, but enforcement is inconsistent. If Deezer relies on self-reported metadata, creators who don't disclose AI involvement won't be filtered; those who do will face a discoverability penalty.
This creates a perverse incentive: honest disclosure hurts reach, silence doesn't. Until there's a reliable detection layer — something like Google's SynthID approach applied to audio — the filter is only as good as the honesty of whoever uploaded the track.
For AI musicians who are transparent about their process, the near-term calculus gets harder. A producer using tools like Suno or Udio to generate backing tracks and then layering live vocals faces a binary label that doesn't capture that nuance.
What this signals for the broader distribution landscape
Deezer's move doesn't happen in a vacuum. The DistroKid-UMG lawsuit over AI-generated content flooding streaming catalogs has put every distributor on notice that inaction carries legal risk. Deezer's filter is partly a listener experience feature and partly a liability hedge — showing regulators and rights holders that the platform is taking AI content seriously without actually removing it.
Spotify, Apple Music, and Amazon Music have not announced equivalent features, which means Deezer is running a real-world test of listener demand. If engagement metrics show that a meaningful portion of users enable the filter, expect the others to follow within a product cycle or two.
For creators building an audience on streaming right now, the practical move is to get ahead of the metadata question: check what your distributor reports in the AI-content field for your releases, and make a deliberate choice about disclosure before the filter goes live. Once Deezer's toggle is on and your tracks are labeled — or not — changing that retroactively may not be straightforward.
The filter rolls out before year's end. That's a short runway to get your catalog's metadata in order.
