Deezer launches a detector for AI-generated music

Streaming service Deezer launched a tool in 2026 that lets users scan their playlists on other music platforms to find tracks created by artificial intelligence. The information appeared on the company's website, deezer.com; the service is aimed primarily at listeners who want to understand where the music in their playlists comes from, regardless of where those playlists were built.

What the new detector does

The new tool differs from the usual AI-content labeling features in that it is not limited to Deezer's own platform and can analyze playlists built on third-party services. This addresses a real problem: a listener may spend years building playlists in one app while using Deezer only as an additional service, and until now they had no way to check how many tracks in their own library were actually generated by a neural network rather than recorded by live performers.

The problem is especially acute for algorithmically assembled playlists, such as mood or background-music playlists, which are generated by recommendation systems rather than a human curator. Anonymous AI tracks most often end up in these playlists: they are well optimized for the key metrics of recommendation algorithms, such as a steady tempo and predictable structure, but they are faceless and not tied to a specific artist whose name a user could learn and whom they could notice.

Why streaming services have taken on the problem of AI music

Over the past few years, the flow of tracks created wholly or partly by generative models has become a noticeable item in streaming catalogs. It ranges from short instrumental tracks assembled algorithmically for specific mood playlists to full songs with synthesized vocals imitating the style of well-known artists. For the industry, this creates several problems at once: royalties get diluted between real artists and anonymous AI channels, bots may inflate play counts to promote AI tracks, and listeners increasingly distrust what they are actually hearing. Labeling and detection tools are the platforms' response to growing pressure from musicians, labels and regulators demanding transparency about where content comes from.

Key facts:

• Deezer is a French music streaming service
• The new feature scans a user's playlists on other platforms
• The goal is to detect tracks generated by artificial intelligence
• The information was published on the official website, deezer.com

What this means for listeners and artists

For the average user, the new tool is essentially an independent audit of their own music library: instead of relying on the labeling of the platform where the playlist was built, they can check it against a separate service with its own method for recognizing AI content. For artists and labels, such tools are gradually becoming part of the industry's trust infrastructure, much as anti-plagiarism systems became standard in text and academic fields. The more platforms independently verify where tracks come from, the harder it is for anonymous AI channels to pass off generated content as the work of live performers and collect the same share of revenue as real musicians.

For the music industry itself, the ability to tell AI content from the work of live musicians is gradually becoming a question not only of ethics but of money: streaming royalties are distributed in proportion to plays, and every AI track that is played while disguised as an ordinary recording takes a part of the shared payout pool away from real performers. Tools like Deezer's detector give the industry a way to measure the scale of this problem quantitatively, rather than relying only on individual high-profile stories about viral AI hits. In time, such detectors may become as familiar a part of streaming services as play counters or recommendation algorithms, only working in the opposite direction: protecting the transparency of the catalog rather than expanding it.

For Deezer itself, launching such a tool is also a way to stand out from competitors, which for now mostly focus on labeling content within their own platforms and do not offer users a way to check the origin of music beyond them.