The music industry spent most of 2025 arguing about whether AI music was a threat. Warner Music Group's CEO Robert Kyncl spent March 3rd telling shareholders it's an opportunity.
In his annual letter to investors, Kyncl described AI as enabling fans to "reimagine" music rather than just listen to it. WMG's CFO Armin Zerza had already told investors in January that the company's Suno partnership is expected to deliver material top and bottom line growth starting in fiscal 2027 — that's October 2026. This is not a hedge. This is a bet.
The timing is striking. Apple Music announced AI Transparency Tags just days ago, framing the conversation around disclosure and consumer protection. Meanwhile the CEO of one of the three major labels is telling shareholders AI is where the next growth comes from.
Both things can be true. The labels want protection from uncontrolled AI disruption while simultaneously profiting from controlled AI partnerships. That's not hypocrisy, it's strategy. WMG has signed deals with Suno, Udio, Stability AI and Klay. UMG and Sony, by contrast, remain in active litigation against Suno over copyright infringement in model training.
THE NUMBER THAT SHOULD CHANGE THE CONVERSATION
On March 6th, Universal Music Group reported its full-year 2025 results. Buried in the earnings call was a figure from UMG's Chief Digital Officer Michael Nash that reframes the entire disclosure debate. Despite enormous volumes of AI uploads flooding streaming platforms, Nash reported that organic consumption of AI content by actual listeners is less than 0.5% of total streams, based on the best available data.
The top 10 AI acts of 2025 ranked between position 7,049 and 92,141 globally. The most streamed AI act of the entire year barely cracked the top 7,000.
Nash called AI music "an insignificant and comprehensively mitigated risk" to UMG's revenue model. Real listeners are not choosing AI music in meaningful numbers — not because the music is bad, but because discovery is broken. There is no reliable way to find the good stuff in a sea of one-prompt slop.
That's the actual problem. Not disclosure. Discovery.
WHAT THE DATA SAYS ABOUT BIAS
When UMG's own data shows that AI music accounts for less than 0.5% of organic streams, the argument that listeners need protection from unknowingly consuming AI music starts to look thin. They're barely encountering it at all.
What research does show consistently is that labeling affects perceived quality before a single note plays. Tell someone a track is AI-generated and their rating drops, regardless of what they're actually hearing. The label does the work before the music gets a chance.
This is why blind rating exists. Not to hide AI music from listeners, but to give it the same starting position as everything else. Judge the sound, then get the context.
THE REAL DIVIDE
The music industry in 2026 is splitting into two camps. Labels like WMG and UMG that have signed AI licensing deals and see revenue potential. And the artist community watching their session work, sync placements and studio bookings decline while the labels negotiate on their behalf.
Kyncl's letter notes that just 27% of US music streaming consumption in 2024 came from new releases, down from 45% a decade ago. Catalog is king. AI makes generating catalog-adjacent content cheaper than ever. The labels have figured out how to profit from that. The working musicians haven't.
This is the conversation Apple's transparency tags don't address. They focus on the listener's right to know. They say nothing about the creator's right to a fair market.
WHERE THIS LEAVES INDEPENDENT AI CREATORS
If organic AI streams are under 0.5% despite millions of uploads, the problem isn't oversaturation of the listener experience. It's that 99.5% of AI music is invisible, buried under algorithmic bias toward established artists and platform economics that favor volume over quality.
The bot farm fraud problem is real — UMG has had to negotiate anti-dilution clauses into its DSP deals specifically to protect artist royalties from being siphoned by fake streams. But that fraud is the work of bad actors gaming royalty pools, not the independent creators putting genuine effort into their music.
Those creators aren't competing with human artists for listeners. They're competing with the noise floor.
That's the problem VoteMyAI was built to solve. Community-driven blind ratings that surface quality regardless of who made it or what tool they used. No algorithmic bias, no follower count advantage, no label backing required. Just the audio and an honest signal.
The industry data is starting to confirm what we've seen in our own ratings: AI music can be good. The problem is infrastructure. Nobody has built a reliable way to find it.
We're working on that. votemyai.com
HEAR IT FOR YOURSELF
Over 3,900 ratings from real listeners who had no idea what tool made what track. No labels, no bias — just the music.
Explore VoteMyAI →