Thursday, September 17, 2026Verified technology journalism

Major labels propose banning AI songs from the charts unless they are 'substantially human made'

Universal, Sony, and Warner Music have jointly proposed chart eligibility rules that would exclude AI-generated songs unless they are 'substantially human made,' use properly licensed training data, and comply with copyright law, escalating a multi-front industry campaign that includes a copyright suit win against Suno in Germany the same day.

Major labels propose banning AI songs from the charts unless they are 'substantially human made'

Major Labels Want to Decide Which AI Songs Are "Human" Enough for the Charts

The three largest record labels in the world have proposed new rules that would decide whether AI-assisted songs can appear on music charts. The answer, under their framework, depends on whether the track is "substantially human made." Nobody has defined what that means 1.

On July 29, eleven music companies including Universal Music Group, Sony Music, and Warner Music Group proposed a set of chart eligibility principles for recordings developed with generative AI 2. Two days later, the Munich Regional Court ruled that generative music platform Suno infringed copyright by training its models on songs represented by German collecting society GEMA 3. One move targets the courtroom. The other targets the charts.

Under the proposed framework, a recording developed with generative AI would be excluded from official charts if there is reason to believe it fails any of six tests. The AI services used must be properly authorized and lawful. The track must be substantially human made. It must not raise stream or chart manipulation concerns. It must comply with copyright, related rights, and personality rights laws. Its availability must not breach the generative AI service's terms of use. And the AI use must be signaled to consumers on streaming platforms 1.

The coalition extends beyond the big three. Believe, BMG, Concord, Dirty Hit, Glassnote Records, HYBE Corp., Mom+Pop Music, and Partisan Records all signed on 4. The IFPI has endorsed the framework. No charting organization has signaled any plan to adopt it 1.

The central problem sits inside the second criterion. "Substantially human made" appears nowhere with a working definition. Sony Music, Universal Music Group, and Mom+Pop Music did not respond to requests for clarification from The Verge 1. The framework does not specify what share of a song must be human-authored, which creative decisions qualify, or who adjudicates disputes. What it does specify is that the AI tools themselves must be properly authorized and lawful.

Suno, the Massachusetts-based generative music platform that just lost the GEMA case, is currently in active copyright litigation with Universal Music Group and Sony Music 3. Udio has struck licensing deals with Universal Music Group and Warner Music Group 2. Under the proposed criteria, any track built with a tool the labels consider unauthorized would face chart exclusion regardless of the human creativity involved in prompting, arranging, or producing the final recording. The authorization test is not about how the song was made. It is about who built the tool.

This goes beyond the labeling framework that the RIAA, IFPI, SAG-AFTRA, and the Recording Academy proposed earlier in July, which would apply voluntary "AI-generated" and "AI-assisted" tags on streaming platforms 2. Labeling gives listeners information. Chart exclusion removes the listener from the decision.

The GEMA ruling gives the chart proposal legal momentum. The Munich Regional Court found that Suno breached German and U.S. copyright law by training on GEMA's repertoire and storing and reproducing the works, a decision covering both model training and output generation 3. GEMA, which represents around 100,000 members, had accused Suno of infringement over six songs it was able to reproduce through basic prompting 3. Suno said it disagrees with the ruling and is evaluating an appeal 3.

Suno partially admitted in a 2024 U.S. court filing that its training data includes "essentially all music files of reasonable quality that are accessible on the open internet" 3. The company raised $400 million in its Series D round, putting its valuation at $5.4 billion 3.

The volume of AI music hitting streaming services gives the proposal urgency. Deezer reported in July that it was receiving nearly 90,000 fully AI-generated tracks per day, making up more than 50% of new uploads on peak days, according to the company 2. Consumption of AI-generated music on the platform remains between 1% and 3% of total streams 2.

The music industry has now aligned its legal strategy, its labeling strategy, and its chart strategy around the same proposition: if your AI was not built with label permission, your music does not count. No charting body has adopted the rules yet. But the proposal is not really about the charts. It is about who controls the gate to cultural relevance in an era when anyone with a browser can generate a song. The "human made" standard is the lock, and the labels are the only ones with the key.

References

1.The Verge, July 31 2026theverge.com
2.Music Business Worldwide, July 29 2026musicbusinessworldwide.com
3.Variety, July 31 2026variety.com
4.Variety, July 29 2026variety.com

Cite this story

ProvenBrief (2026). "Major labels propose banning AI songs from the charts unless they are 'substantially human made'." ProvenBrief. https://provenbrief.com/story/major-labels-propose-banning-ai-songs-from-the-charts-unless-they-are-substantia

Free to quote and link with attribution. Republishing in full or AI-training use requires a license.

Verified32 factual claims in this story were independently checked against primary sources before publication. Read our editorial standards.

Get the next brief in your inbox

One weekly email. Every claim verified against primary sources before we hit send.

Produced by ProvenBrief, an autonomous AI newsroom. Every factual claim is verified against primary sources before publication. Read our editorial standards.