Written by Many readers will already have seen the headlines. Generative AI is flooding the streaming platforms, on Deezer alone, AI-generated tracks now account for as much as half of all daily uploads on peak days, tens of thousands of fully synthetic songs arriving every single day, worldwide, without pause. I had a longer piece on what that actually means for musicians nearly finished this week when Suno announced a new policy that stopped me in my tracks, not because it changes anything I’d already written, but because it’s such a clean example of exactly the pattern I’ve spent weeks unpacking that I didn’t want to let it pass by as just another news item.
Here’s what happened. From September 3, Suno is capping how much users can download off its platform. Free accounts get seven downloads, total, for the lifetime of the account, applied retroactively to songs already made. Paid subscribers on the $8 to $10 a month Pro tier get 20 downloads a month with commercial rights attached. Premier subscribers, paying $24 to $30, get 60 a month. Suno’s own framing is that this protects a healthy music ecosystem by making it harder for bad actors to mass-export low-effort AI content, the stuff most of us have started calling slop, off the platform and out into the wild.
I’ve been reading enough of these announcements lately to know what to check first before taking that framing at face value, and it doesn’t hold up especially well. Streaming and generating music inside Suno remain completely unlimited on every single tier. And subscribers who use Suno Studio, the company’s professional production suite, face no download limit whatsoever. So the cap lands almost entirely on casual users and free accounts, while the tier best suited to actual industrial-scale output is explicitly carved out. A genuine content farm, the kind pushing tens of thousands of tracks a day into streaming platforms, loses close to nothing here. If this were really built to stop bulk exporters, the exemption would run the other way.
So who does it actually protect? Follow the money rather than the press release and a much clearer picture emerges. Suno settled with Warner Music Group back in November, and the two are now, in effect, commercial partners, Warner has a direct financial stake in Suno looking like a responsible, legitimate operator. Universal and Sony, meanwhile, are still suing Suno in the United States, arguing its models were trained on their catalogues without a licence. A German court sided against Suno on almost exactly that question last month, ruling it had infringed copyright by training on songs represented by the collecting society GEMA without permission. Independent musicians have their own separate class action running against the company too. Nobody in this story is a disinterested party quietly looking out for musicians. Warner benefits from Suno appearing legitimate. Suno benefits from appearing responsible while it’s mid-lawsuit. Universal and Sony benefit from Suno looking reckless. A download cap that mostly inconveniences casual users while exempting the highest-volume tier reads a great deal more like litigation strategy and reputation management than genuine reform, and the fact that it leaves the real export mechanism almost untouched is the tell.
I don’t think this pattern is unique to Suno. The same dynamic, an interested party dressing up self-protection as ecosystem stewardship, is turning up right across this industry at the moment, in how labels are proposing new rules, in how platforms are rolling out new badges, in how coalitions are announcing standards they can’t yet agree on internally. And that’s before you even get to what’s actually happening inside the streaming royalty pool itself, which is a different fight entirely, and arguably the bigger one.
I’d already decided, before any of this landed in my inbox, that I wanted to start writing regularly on the business behind the music, not just the news of the week, but the actual mechanics underneath it, the kind of thing fifty years on both sides of this industry gives you a fairly sharp eye for. The Suno story simply turned out to be the first subject that deserved the full treatment, and the timing worked out well enough that I couldn’t resist opening with it. I’ve spent the last few weeks pulling that whole picture apart properly, who genuinely has the most to lose from AI’s arrival in music, what the data says about what listeners actually want, and just how much of the industry’s response so far deserves to be trusted. It’s a three-part series, and it’s only the beginning of what I’ve got planned for this column. The first instalment goes up shortly. Watch this space.
Last modified: August 12, 2026
