Yandex Music: 37 percent of new releases in first half of 2026 turned out to be AI-generated music

A Habr author who published an investigation into AI-generated music on Yandex Music one and a half months ago has presented a new algorithm for detecting AI tracks with accuracy approaching 100% and analyzed Yandex Music releases for the first half of 2026.

What the new analysis revealed

In the first six months of 2026, 37% of all new releases on Yandex Music show signs of neural network generation. If we consider only tracks that received likes from listeners, the share of AI music rises to 50%—meaning every second popular release could have been created by a neural network rather than a human.

• The author's first article about neural slop on Yandex Music was published one and a half months ago
• In the first half of 2026, 37% of new releases on the service contain signs of AI generation
• Among releases with likes, the share of AI music reaches 50%
• The new algorithm detects generated tracks with accuracy approaching 100%
• GPU is not required for the algorithm to work

How the picture changed over one and a half months

In his first article, the author estimated the share of neural slop on Yandex Music at approximately one-third of the catalog. The new study, built on a more accurate algorithm and covering all releases from the first half of 2026, shows not just a comparable figure of 37%, but a more alarming signal: among tracks that listeners actually enjoy and that receive likes, AI music now comprises half. This means that neural tracks are not simply filling the catalog by default, but competing with human-made music for audience attention and appearing in playlists on equal terms.

Why detection works without GPU

Most modern systems for recognizing AI content rely on neural network models that require video cards for inference. The algorithm author emphasizes that his approach is structured differently and does not use GPU—this makes verification cheaper and more accessible for mass scanning of streaming service catalogs, where the count of releases runs into hundreds of thousands of tracks. The absence of expensive hardware requirements is a key condition for applying such detection at industrial scale rather than just within a single research project.

What this means for charts and royalties

If half of the releases with likes are generated by neural networks, this poses a question for the platform about how to distribute audience attention and royalties between live performers and generative projects. So far, Yandex Music has not commented on the author's estimates and has not reported on its own tools for marking AI tracks in the catalog, which means listeners currently cannot distinguish neural music from tracks by live performers without a third-party detector.

What this means

If independent AI music detection with accuracy approaching 100% truly works without GPU, streaming services and rights holders get a cheap tool for marking or filtering AI-generated content—the question is whether platforms like Yandex Music will actually apply it in practice.

Frequently Asked Questions

How much AI music is on Yandex Music in 2026?

According to the author's analysis, in the first half of 2026, 37% of all new releases on the service show signs of AI generation, and among releases with likes, this share reaches 50%.

Is GPU needed to detect AI music?

No—the algorithm author emphasizes that the detector does not require a video card, unlike many other ML approaches to content recognition.