This week in AI music: Qobuz and Deezer clarify their stances

At the start of the year, as part of my music resolutions piece, I proposed a rejection of AI Music.

I previously mentioned how Spotify and other streaming services are dealing with AI music on their platforms, as Bandcamp put out a strong statement banning AI music from its platform, other leading platforms are now followign suit in sharing their stances.

---

Qobuz

Qobuz have published an AI charter with the top line of reiterating their previous stance and clarifying how AI is used won Qobuz.:

The heart of Qobuz is and will remain human: editorial curation, music expertise, content creation.

We use AI to enhance user experience and optimize how we work. We refuse to let it replace human judgment, artistic sensibility, or the relationship with our subscribers.

It also says it has a zero-tolerance policy for hosting AI-generated content, and will remove songs and releases when identified.

We are strongly committed to fighting fraud and apply a zero-tolerance policy when it comes to AI-generated content and AI-driven streaming activities. Our measures include:

• Detection and monitoring systems for AI-generated content (in development) and fraudulent streaming patterns (effective)
• Right to remove fraudulent catalogs when issues are identified
• Identification of fraudulent streams, which are excluded from reporting and royalty payments
• Contractual clauses prohibiting delivery of 100% AI-generated content
• Publication of our AI charter on the website for the end-users
• Regular workshops to increase employee awareness

Qobuz’s redlines:

We never generate:

• Musical content to feed our catalog or calculate royalties

We never replace humans:

• No replacement of editorial curation, artistic validation, or creative decisions
• No manipulation of our recommendations for commercial gain
• No promotion of fraudulent or 100% AI gen content

How Qobuz uses AI:

• We use AI to automate low-value tasks and improve service efficiency:
  • Limited technical support: transcription, content moderation, first-level customer support
• • Enhanced user experience: improved search
• • Internal operations optimization: security, automation of repetitive processes
  In all cases, systematic human validation is mandatory. Our teams analyze, control, and assume final responsibility for any AI-assisted production.

In doing so, it also acknowledges limitations and grey areas when grappling with an evolving technology.

It all seems pretty sensible to me, however the fully-generated and well-known AI artist Xania Monet is still present on the platform as one example I quickly search.

You can read the full charter here.

---

Deezer

Deezer have announced their proactive measures to address AI, as it says it is receiving over 60,000 fully AI-generated tracks every day – accounting for more than 39% of the total daily delivery – it was previously reported as 18%.

These include selling their AI music detection tool to allow other streaming services to do the same.

Deezer confirms demonetization of up to 85% of AI-music streams due to fraud and moves to sell AI-detection Technology

• A year after the launch of Deezer’s AI-music detection tool, the company is now making it commercially available, and encourages industry wide transparency.
• Over 13.4M AI-tracks have been detected and tagged on Deezer in 2025, and over 60,000 AI-tracks are now uploaded per day, equal to roughly 39% of daily intake
• Up to 85% of all streams on AI-generated music are detected as fraudulent – they are demonetized and removed from the royalty pool, as Deezer continues to support fair payments for artists and songwriters

All fully AI-generated songs are automatically removed from algorithmic recommendations and are not included in editorial playlists.