YouTube Music Starts Labeling AI Songs

"Our goal is to make it as easy as possible for viewers and listeners to have the right information and clarity when it comes to the music they listen to."

Date:

Listeners on YouTube Music will soon know when a song wasn’t made by a human. Starting this week, the platform is labeling tracks created primarily with generative AI.

What We Know: YouTube revealed the change Wednesday on its official blog. The label sits on a track’s player page. Listeners can tap it to see how AI contributed to the song. Eventually, the labels will expand to the main YouTube app as well.

- Advertisement -

The labels target songs where AI played a significant role. For example, a track that used a generative model for both the composition and the vocals would qualify. To find those songs, YouTube is combining input from music partners with its own detection tools. Partners disclose AI use through three categories: Fully Gen AI, Partly Gen AI and No Gen AI. Vocals touched up with pitch-correction tools like Auto-Tune still land in the No Gen AI bucket.

However, YouTube isn’t penalizing AI music. According to TechCrunch, AI involvement alone doesn’t make a track ineligible for monetization. That approach differs from the brand’s notable rivals. Deezer, for instance, keeps detected AI tracks out of algorithmic recommendations and editorial playlists. The service also reported receiving roughly 90,000 fully AI-generated tracks every day.

Meanwhile, other platforms are moving in the same direction. Spotify already uses AI labels and a spam filter. Additionally, Apple Music plans “Made With AI” labels later this year. TIDAL, by comparison, lets listeners switch off fully AI-generated recordings. Ironically, YouTube also experiments with AI music itself. Its Dream Track tool lets select creators use AI versions of participating artists’ voices.

What They Said: “Our goal is to make it as easy as possible for viewers and listeners to have the right information and clarity when it comes to the music they listen to. – YouTube

What Remains Unclear: YouTube didn’t say when the labels will reach the main YouTube app. It also didn’t explain how its detection tools work or how accurate they are. Additionally, it is unclear if listeners will eventually receive the option to filter out AI music completely.

What It Means: For the music business, labels are a first step. They are not a solution though. Artists already compete for attention against a growing flood of synthetic tracks. Clear disclosure helps listeners decide what they value. For radio, the shift carries a lesson. Programmers and personalities have long been trusted as human curators. As streaming fills with AI content, that human filter could become even more valuable.

For more on YouTube’s goals and plans for innovation, check out Oprah Winfrey’s conversation with CEO Neal Mohan.

Barrett Media produces daily content on the music, news, and sports media industries. Sign up for our newsletters to stay updated and get the latest information right in your inbox.

- Advertisement -
Jason Barrett
Jason Barretthttps://barrettmedia.com
Jason Barrett is the Founder and CEO of Barrett Media. The company launched in September 2015 and has provided consulting services to America's top audio and video brands, while simultaneously covering the media industry at BarrettMedia.com, becoming a daily destination for media professionals. Prior to Barrett Media, Jason built and programmed 95.7 The Game in San Francisco, and 101 ESPN in St. Louis. He was also the first sports programmer for SportsTalk 950 in Philadelphia, which later became 97.5 The Fanatic. Barrett also led 590 The Fan KFNS in St. Louis, and ESPN 1340/1390 in Poughkeepsie, NY, and worked on-air and behind the scenes at 101.5 WPDH, WTBQ 1110AM, and WPYX 106.5. He also spent two years at ESPN Radio in Bristol, CT producing 'The Dan Patrick Show' and 'GameNight'. JB can be reached on Twitter @SportsRadioPD or by email at Jason@BarrettMedia.com.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Next Article