This won’t be a revelation to regular Music Ally readers: there is a lot of AI music news in 2025. Sometimes, rounding it up into one article offers a useful snapshot of What’s Going On in this space. Which is why we’re doing exactly that today.
First up, some bad news. A large group of music and creative-industry bodies has strongly criticised the latest draft of the European Union’s ‘General-Purpose AI (GPAI) Code of Practice’. They say it “sets the bar so low as to provide no meaningful assistance for authors, performers, and other right holders to exercise or enforce their rights”.
The full statement from the organisations is here, and the IFPI’s crisp summary here.
“Despite having participated constructively in the drafting process and providing substantive comments on previous drafts to address critical issues, the coalition’s key concerns have been ignored,” it said.
“No Code is better than the fundamentally flawed third draft,” added CEO Victoria Oakley.
The second story today is some new data from a company called AITools, which analyses web traffic data for more than 10.5k AI tools. It tracked more than 101.12bn visits to those tools in 2024 – and some of them were musical.
Particularly notable is that according to this data, Suno accounted for 55% of all visits to ‘music generators’ in 2024. That’s 590.3m total visits by 143.7 million unique users, well ahead of second-ranked BandLab (185.7m visits from 50.6 million unique users).
Today’s third story is a white paper from AI-music startup Daaci which should be of keen interest to the music industry.
It’s described as an “ethical, legal framework for creating generative original music, that embeds copyright, consent and attribution for artists at its core” which “shows how sample companies, creators, rights holders and artists can be rewarded to a deep level of attribution”.
This means tracing back from a musical AI’s outputs (the tracks it generates) to understand which inputs (the songs the AI was trained on) contributed to each, which in turn enables royalties to be paid to their creators. Something music rightsholders are understandably keen on seeing happen.
Today’s fourth story is the return of the AI Song Contest, the pioneering competition that showcases how clever, creative humans can use clever, creative AI models to make original music.
The event will take place in Amsterdam this November, with submissions opening in July. There are two new partners this year – Musical AI and RipX DAW – and a strong focus on ethics.
When teams enter the contest, they’ll have to describe the data used to train their models – with potential disqualification for entries that don’t do this. Last year, the contest’s organisers also strongly recommended against using Suno and Udio, after labels sued those startups for copyright infringement.
Finally today, music-tech veteran Virginie Berger of MatchTunes has published her latest guest column for Music Ally, offering her views on how “agentic AI and quantum computing are crashing into the music industry faster than anyone is ready for”.
She runs through some of the tools and models that she has been testing out, and explains why the combination of quantum computing and increasingly-autonomous AI will “inevitably disrupt every facet of the music industry, from how music is distributed and protected, to how it is created and monetised”. You can read her full column here.