I don’t consider myself adverse to technology, nor do I think people who know me would describe me like that. I have been fascinated by machines and technology since I was very young and I still am. Nonetheless, expressing hesitancy or frustration about the spread of generative AI or the proliferation of data centres can be enough to get you labelled a luddite, or if you’re unlucky, something far more unhinged.
Last month, I published an article about music and the AI industry here on my blog. Among other things, I discussed the German performing rights organisation GEMA suing the American generative AI company SUNO Inc., creator of the music generator app Suno AI. As stated in a press release by the law firm Härting which represents GEMA, the organisation had demonstrated the ability to use Suno AI to reproduce several popular songs protected by GEMA. Thus, they allege that the generative AI tool had been illegitimately trained on copyrighted songs, including those protected by GEMA.
Strength in Numbers
Since writing that article, I have learned that the Danish performing rights organisation Koda is also suing SUNO Inc. In a series of articles on their own website, Koda presents a detailed overview of their case against Suno, including multiple egregious examples of Suno AI’s capacity to reproduce copyrighted songs. The clearest, most educational presentation is on Koda’s AI landing page: a video comparing the first bars of Suno AI’s uncanny simulacrum of 1990s earworm Barbie Girl against the original.
Koda appears to be facing the threat of the AI industry head-on. In July last year, Koda wrote on their website about a then just-released joint statement by a large number of rights holders across Europe put out by the central rights organisation CISAC, criticising the European Commission’s AI Act.
Mette Vestergaard Andersen, Koda’s Director of Business Development, is quoted on their website stating that rights holders “are left with an implementation package [of the AI Act] that offers more advantages to AI providers than to the rightsholders the AI Regulation was supposed to protect”. While acknowledging “the good intentions of the AI Regulation”, decries the AI Act’s “worryingly low” level of ambition and that it “risks opening the floodgates for AI services to exploit the work of musicians and artists without payment or permission”.
See You in Court
I mentioned the AI Act in my article last month alongside the United Kingdom’s controversial Data Use and Access Bill. The UK government refused an amendment to the bill which would have required tech companies to declare their use of copyrighted material when training AI tools.
The reason for their refusal was, as described in the BBC article linked above, that “it would stifle the AI industry and result in the UK getting left behind in this lucrative and booming sector”. I thought the EU’s AI Act looked like it offered plenty of exceptions and loopholes for tech companies to abuse – completely legally – much like the UK bill. I feel a bitter satisfaction that much more competent people than me, i.e. the representatives of CISAC, feel the same way.
Earlier this year, SUNO Inc. actually admitted in a court hearing in the case brought by GEMA that it had indeed trained its generative AI model on copyrighted music. The company argued, however, that it fell under the U.S. legal doctrine of “fair use”. Very briefly, “fair use” is supposed to allow exceptions for use of copyrighted material for satirical, academic and similar purposes. SUNO Inc., on the other hand, has used copyrighted material to develop a commercial service – one that on top of that threatens the livelihood of those creators whose works they have ripped off.
I haven’t been able to find any information about Koda’s case against SUNO Inc. The Munich court’s decision in GEMA’s case against SUNO Inc. was supposed to come on June 12th, but in late May that date was postponed more than a month until Friday, July 31st. Last November, GEMA won a court case against OpenAI, which they described as a “landmark ruling against AI”. I assume GEMA hopes that win will be viewed as precedent in this new case, and I imagine that would also steel Koda’s spine. I for one await the court’s decition with bated breath.
Going Toe to Toe
Koda in Denmark, along with GEMA in Germany, is actively taking the fight to the AI industry. The other Nordic rights organisations seem to handle things in different ways. The Norwegian organisation Tono simultaneously posted their own article highlighting the joint statement from rights holders across Europe, published centrally by the international collective rights organisation CISAC. Tono published another article a few weeks before about lobbying in the European Commission on behalf of creators’ rights.
The Finnish and Swedish organisations, Teosto and Stim respectively, appear comparably placid and docile. Teosto wrote in March 2025 about being “committed to working with AI companies [to create] a balanced, sustainable ecosystem where both rights holders and technology innvators can thrive”. Both Teosto and Stim posted on April 29th 2025 about all five Nordic rights organisations agreeing on a joint set of principles for licensing music to AI companies. Other than that, there’s been a dearth of AI-related updates from either of them.
The Icelandic rights organisation Stef also wrote about the Nordic organisations’ joint principles, as well as highlighting in a separate article CISACs late-2024-published study about the potentially disastrous economic impact of generative AI on human creators. Back in November 2025, Stef also brought attention to both Koda’s lawsuit against SUNO Inc. and GEMA’s court victory against OpenAI.
A side note about Koda is that they published a report in September 2025 that they had commissioned about how the Danish music industry specifically would be impacted by generative AI. The analysis, which came to similar depressing conclusions as CISAC the year before, is available to read from Koda’s website.
In addition, as recently as this February, Stef wrote about a survey on the impact of music in store environments conducted in 2025 by its Austrian counterpart, AKM. Among the results of the survery Stef highlighted was a broad disregard for AI-generated music among both staff and customers. The original study in German can be found in this article on AKM’s website. Stef also hosts an English version of the study.
Low Expectations
I of course don’t know what neither Stef, Tono, Teosto, nor Stim actually does to combat the rise of generative AI and its effects on us creators. Nor do I think simply posturing aggressively or highlighting the work of other organisations is enough to make an actual difference. But at the very least, Tono and Stef appear to be more decisively acting in favour of us creators and rights holders, which is more than I can say about Stim and Teosto, which frustrates me.
Stim recently held a follow-up to its livestreamed presentation back in May about its perspective music and AI. As I discussed in my previous article, that first presentation in May worried me and even started to erode my trust in Stim’s ability – even its desire – to actually fight to preserve my rights. This second presentation only made me feel even more despondent.
Stim’s representatives talked about how they preferred negotiating licenses with AI companies instead of waiting for regulation from lawmakers, because licenses “give us much more control” and because “regulation take a lot of time to work out”. This is a very confusing stance for Stim to take, considering that CISAC, which Stim is effectively a part of, explicitly called out the European Commission in its joint statement to both reinforce and enforce EU regulation!
Also, perhaps to reinforce their licensing-before-regulation position, Stim’s representatives showed how the world’s largest music publishers in Autumn 2025 begun making licensing deals with AI companies such as UDIO, KLAY and SUNO. However, the representatives neglected to highlight the very important fact that these licenses only came after Sony, Universal and Warner all sued SUNO and UDIO. In fact, these license agreements seem to have been part of settlement deals with Universal and Warner. Sony, meanwhile, is still fighting both SUNO and UDIO – in two different U.S. courts.
Towards the end of the presentation, Stim’s representatives took a few questions from the viewers. In one of their answers, they again pushed for licenses, arguing that licenses would give companies access to very desirable, much higher-quality data than illegal scraping and stream ripping has given them thus far. But they immediately followed that up with admitting that the companies “learned a hell of a lot by behaving like shit to begin with”. In other words, they seem aware of the fact that AI companies have behaved poorly – even provably illegally – but also very hesitant to admit that fact openly, instead striking what feels to me like an almost submissive, or at least passive, tone.
