Machine-made songs are 44% of Deezer's daily uploads but just 1% of Spotify plays. As they flood a fixed payout pool, real songwriters could lose nearly $1 billion a year — a threat the market thinks is already solved.
Nearly half of the tracks uploaded to Deezer each day — 44% — are made entirely by machines, while only about 1% of Spotify's plays go to such songs today. That gap is the threat. Streaming pays from a fixed pool split across every track, so flooding it with cheap machine-made music means each real song earns less. Our model puts the yearly loss to United States songwriters at $992 million, about 5.6% of the $3.5 billion pool, and an independent forecaster sees a 24% hit to creator income by 2028.
The market underprices this because public attention to tools like Suno has collapsed roughly 97% from its peak, making the problem feel solved even as supply keeps rising.
Spotify Technology runs the platform absorbing the flood and removed 75 million spam tracks in a year. Universal Music Group and Warner Music Group own catalogues whose per-play value falls as the pool dilutes. Reservoir Media and Believe, smaller rights holders, have thinner cushions against the same squeeze.
Why this matters. Machine-made songs now make up nearly half of daily uploads on one major service, and they cost almost nothing to produce. They split a fixed pool of listener money among far more tracks, so each real song earns less. Anyone who lends to, runs, or invests in music-rights owners and streaming services should care, because the income those royalties throw off could quietly shrink.
Blindside · Creator Economy
Machine-Made Songs Flood Streaming
Computer-made tracks quietly drain pay from human musicians
Building
80
Blindside index
What drives it — drag to test
each slider starts at our cited estimate — drag to see the range
Share of listening that goes to machine-made tracks10%
Sourced — about 1% of Spotify plays today; one forecaster sees 20% of streaming money by 2028.
Share that takes pay from a real musician45%
Our judgment — no hard proof; machine tracks skew toward background listening, so they may replace less.
Drag on the pay rate from a crowded pool+15%
Our judgment — no hard proof; more tracks split the same pot and push human songs down.
Time to impact
2–4 yearsBuilding
now3 yrs7+ yrs
When the financial hit begins to land, on our read.
How to read this. Drag any slider to test your own number — the chart and index update live. The likelihood and the locked facts stay put.
Yearly pay lost by United States songwriters
$993m5.61% of sector
Dark line = most likely · faint lines = low–high (8 in 10 outcomes land between) · shaded band = what outside analysts expect
Our estimate lands within what outside analysts expect ✓
Chance this is a permanent shift, not a blip
66%
Average of five independent reads (range 50–74%):
The track record68%
A fixed pot is split per play; flood it with supply and each play pays less. The catalogue just passed 250 million tracks.
How it works74%
Recommenders chase listening time, not who made the song. Machine tracks cost almost nothing and are 44% of Deezer's daily uploads.
The skeptic's case50%
Platforms are purging 75 million tracks, machine music is still 1% of plays, and labelling can wall it off early.
What forecasters say72%
An independent forecaster prices a 24% hit to creator income by 2028 as their middle case, not a worst case.
What platforms reveal66%
Spotify shipping spam filters and disclosure rules shows it sees the dilution risk as real and large.
Fixed — the sliders change the size of the hit, not the odds it's permanent.
Why this matters
Machine-made songs now make up nearly half of daily uploads on one major service, and they cost almost nothing to produce. They split a fixed pool of listener money among far more tracks, so each real song earns less. Anyone who lends to, runs, or invests in music-rights owners and streaming services should care, because the income those royalties throw off could quietly shrink.
Most exposed companies
Spotify Technology SPOT · Universal Music Group UMG · Warner Music Group WMG · Reservoir Media RSVR · Believe BLV
🔒
The facts — locked
measured, not editable
44%
44% of tracks uploaded to Deezer each day are entirely machine-made
Deezer · TechCrunch Apr 2026
1%
1% of Spotify plays and payouts come from machine-made music today
Spotify 2025
−24%
24% of music-creator income expected to be eaten away by 2028
CISAC · PMP Strategy 2024
$4.2bn
$4.2 billion in yearly creator income lost worldwide to generative tools by 2028
CISAC (€4bn)
75M
75 million spam or machine-made tracks Spotify removed in 12 months
Spotify Sept 2025
$3.5bn
$3.5 billion United States songwriter performance-royalty pool, the base figure
ASCAP + BMI + SESAC 2023
↓97%
Public attention to the Suno tool fell 97% from its April 2025 peak to May 2026
Wikipedia pageviews (live)
Attention is falling while the impact compounds. the blind spot is widening, not closing.
Our central estimate is $992 million in United States songwriter royalties displaced each year, or 5.6% of the $3.5 billion pool. That matches an independent forecast of a 24% hit to creator income by 2028. Our threat score is 80 out of 100 with a 66% chance this is permanent — a danger the market treats as solved only because public attention to these tools has fallen about 97% from its peak.
