A few weeks ago I had a conversation with a brilliant person from the creative industry. What she said stayed with me, and it pushed me to write this down.
The discussion around how to regulate artificial intelligence in creative content sounds remarkably familiar. When digital first appeared, many wanted to protect the physical business by making downloads more expensive, blocking platforms, and treating anything new as an existential threat. Record labels sued teenagers. Napster was shut down. The industry spent years fighting the inevitable instead of shaping it.
Today the exact same arguments are being recycled. Only now we replace the word "digital" with "AI."
The Arguments We Have Heard Before
"It is going to destroy the industry." That was said about CDs when cassettes appeared. It was said about MP3s when CDs were still selling. It was said about streaming when downloads were the model. Each time, the industry survived. Each time, those who adapted early ended up in a stronger position than those who resisted.
"It cannot be allowed to compete with the current model." This logic led to catastrophic decisions in the digital era. Instead of building Spotify in 2001, the industry tried to stop the future with legal action. By the time streaming arrived with legitimate licensing frameworks, a decade of value had already leaked to piracy and to early tech platforms that moved faster.
"It is not fair how the existing catalog is being used." This is the most legitimate of the three arguments, and it is the one that deserves serious attention. The difference is that fairness requires negotiation, not prohibition.
What Is Actually at Stake
Without meaningful regulation there is a massive transfer of value from the creative world to the technology world. The scale is not trivial. A 2024 CISAC report projected that between 21 and 24 percent of creator incomes in music and audiovisual content could be at risk by 2028. Generative AI models were built on creative work that was not compensated. That is a real problem.
But the other risk is equally real. If regulation is designed primarily to protect incumbent revenue structures rather than to build sustainable frameworks for the future, we could kill the next wave of musical and creative innovation before it even begins. Just as heavy-handed responses to digital could have permanently damaged the music ecosystem we enjoy today.
The challenge for Collective Management Organizations, rights holders, and regulators is to find the path between these two failures.
The Questions That Actually Matter
The debate should not be centered on whether AI should access creative content. That ship has sailed. The real conversation is about the terms.
How should catalog licensing for AI training work? Should it be opt-in, where creators actively choose to make their work available? Should it be a collective licensing mechanism managed by CMOs, as proposed in the United Kingdom by CLA, ALCS, and PLS? Or should there be statutory licensing with mandatory compensation, similar to how radio broadcasting has operated for decades?
How do we capture part of the value generated by AI outputs? If a language model generates a song in the style of a particular artist, or summarizes a body of creative work, some of that generated value should flow back to the creators whose work shaped the model. The attribution technology to make this possible is beginning to emerge, with companies like Sureel AI developing algorithms that can trace which works influenced a given AI output.
On what economic basis should these rights be calculated? Subscription revenue? Advertising? Actual usage per output? These are not simple questions, and they require creative industry stakeholders to be actively involved in designing the frameworks, not just reacting to what technology companies propose.
What Collective Management Organizations Can Do Now
Societies of Collective Management have a structural advantage in this moment. They already represent large catalogs of creators. They already have legal standing to negotiate on behalf of members. They already have the collection and distribution infrastructure that would be needed for any licensing framework at scale.
The passive approach, waiting for governments to regulate or for tech companies to propose frameworks voluntarily, cedes enormous influence. The active approach is to develop clear positions on each of these questions, to engage with AI companies directly, and to build or adopt the technology needed to enforce whatever agreements emerge.
We cannot stop technology. We can only slow it down or redirect it. The question is whether the creative industry wants to shape the redirect or simply observe it.
The Labs at global.esur tracks the most pressing technological challenges in collective management, including AI licensing frameworks and attribution technology. Get in touch if you want to explore what this means for your organization.