YouTube Music has announced labels that let listeners instantly identify songs made using generative AI. The labels will appear on the playback screen, and tapping one will show which parts of the production involved AI. The same feature will also roll out to the main YouTube app.
The player will show AI’s role
What stands out about the new system is that it doesn’t reduce the question to a simple yes or no on whether AI was used. Listeners will be able to see which stages AI contributed to, such as composition or vocals.
The labels will appear on the playback screen, where listeners already look regularly. Rather than hiding the information deep in settings, the design puts it somewhere it can naturally catch the eye while a song is playing.
YouTube says the labels will also expand to its main app, bringing them to people who don’t use the dedicated music app. It’s a fitting move for YouTube, where video and music go hand in hand.
AI involvement organized into three categories
Music partners can choose one of three labels to declare for their work: “Fully Gen AI,” “Partly Gen AI” or “No Gen AI.”
“Fully Gen AI” applies when a song is entirely generated by AI. “Partly Gen AI” refers to works that incorporate AI-generated elements alongside contributions from artists, such as live performances. “No Gen AI” includes songs touched up with pitch-correction tools such as Auto-Tune.
The distinction here is a subtle but important one. Pitch correction is a technology that has been used in studios for decades, and it is treated separately from generative AI. At a time when the term “AI” is often used as a catch-all, this distinction could help prevent listeners from getting the wrong idea.
A system that doesn’t rely solely on self-reporting
YouTube says it will work with music partners and use its own detection tools to identify AI-generated music.
Relying only on self-reporting could give an advantage to those who don’t disclose their use of AI. Using detection tools as well helps lay the groundwork for trustworthy labels. Since the presence or absence of a label can inform listeners’ judgments, the system behind the scenes will be crucial.
Other services opt for exclusion; YouTube opts for disclosure
The influx of AI-generated music is a shared challenge across the industry, but services are responding differently. Apple Music plans to introduce “Made With AI” labels in the second half of this year as part of its AI Transparency Tags. Spotify also has AI labels and music spam filters.
Deezer takes the strictest approach. As of July, it reported receiving nearly 90,000 fully AI-generated tracks a day. It labels the tracks it detects, excludes them from algorithmic recommendations and editorial playlists, and does not pay royalties on fraudulent streams. It is also preparing an option to filter out AI-generated music.
TIDAL also labels recordings it determines to be entirely AI-generated. Listeners can hide them, and they are excluded from royalty distributions.
The choice not to stop monetization
In contrast, YouTube is focusing on disclosure rather than imposing blanket restrictions. The new labels are intended to tell listeners what role AI played; a song is not automatically excluded from monetization simply because AI was involved.
YouTube is also experimenting with AI music creation itself. Dream Track is a tool that lets some creators generate soundtracks using AI recreations of participating artists’ voices.
One approach excludes AI-generated music; the other shows listeners what’s involved and lets them choose. Which is right depends on what listeners want. But as the divide between “AI” and “human” becomes harder to sustain, showing who contributed what could broaden the ways people enjoy music. The value of the system will depend on how self-reporting and detection work in practice, and on how clearly the labels are presented.
Comments