YouTube Music now tags tracks made by AI

That catchy track stuck in your head might have been written, composed, and sung by a neural network before lunch. As synthetic music quietly floods our playlists, YouTube wants listeners to know what they are actually hearing.

YouTube Music officially rolled out new labels to flag songs created primarily using generative AI. The tag appears right on the main player screen, and tapping it pulls up a simple menu showing what role AI played in the production. The feature will also expand to the main YouTube video app soon.

Three tiers of disclosure

Instead of a vague warning, YouTube is teaming up with music industry partners under a three-tiered system: Fully Gen AI, Partly Gen AI, and No Gen AI.

A track built from prompt to finish goes under Fully Gen AI, while live instrumentation mixed with AI stems falls under Partly Gen AI. Crucially, studio staples like Auto-Tune and standard pitch correction fall squarely into the No Gen AI bucket, protecting traditional studio workflows.

If uploaders fail to disclose the AI elements voluntarily, YouTube says it will deploy its own proprietary detection tools and backend signals to catch them.

Clashing platform playbooks

The move arrives during an identity crisis for music streaming. Platforms are picking sides fast, and their approaches could not be more different.

Deezer reported in July that it was absorbing nearly 90,000 fully AI-generated tracks every single day. In response, Deezer strips them from editorial playlists, blocks algorithmic recommendations, withholds royalties from fraudulent streams, and is building a toggle for users to hide them completely. TIDAL takes a similarly aggressive stance, letting users disable AI music and blocking royalty payouts.

Meanwhile, Bandcamp banned AI-generated music outright earlier this year. Spotify limits recommendations for AI personas, and Apple Music is preparing its own transparency tags later this year.

YouTube, however, is not slamming the door. AI-labeled songs remain eligible for monetization and streaming royalties. YouTube is even experimenting on the creative side with Dream Track, which lets creators generate instrumentals with participating artists' voices.

Why it matters

We think YouTube's choice to label rather than ban reflects its identity as an open creator ecosystem rather than a walled digital record store. Drawing a hard line between generative tools and production standards like Auto-Tune is a smart move that avoids alienating working producers.

For everyday listeners, this gives immediate clarity on whether a raw vocal performance came from a human throat or a server rack. If you are an artist, the framework offers some relief that technical assistance will not automatically brand you an imposter.

However, leaving the monetization pipeline wide open creates an obvious risk. When bad actors can still collect streaming payouts, labels alone will not stop bulk uploaders from pumping out cheap AI elevator music to farm royalties. Whether YouTube's internal detection can outsmart coordinated streaming farms remains to be seen.