In the AI gold rush, it’s the artists who are being mined. As millions of machine-made tracks flood streaming platforms, triggering panic over royalties and attribution, creators are left to confront a system being built without them. For many, the real issue isn’t just about lost income; it’s about the fundamental exploitation at its core.
Kartini Ludwig, TEDxSydney speaker and founder of Koup Music and creative studio Kopi Su, is a veteran of the art and tech space. Her background includes work with Google Creative Lab, and projects for the Sydney Opera House. Now, she’s sounding the alarm. For Ludwig, the industry is facing more than a technological threat. It’s an exploitative loop designed to profit from the very people it’s displacing.
An insulting loop: It's not just that Silicon Valley’s AI powerhouses have scraped the internet's creative commons without permission; it's the audacity of what they do next that reveals the true nature of their actions. “I speak to artists who are just beginning to understand that their data has been scraped,” Ludwig says. “But the part they're still wrapping their heads around is that their own work is then being sold back to them. It’s one thing to have your data become part of a tool you never gave consent to; but it’s another to then be asked to pay someone else for it.”
Napster déjà vu: “When we look back at the Napster era, and how destructive some of those deals were between streaming platforms and labels, we can see that history potentially repeating itself right now,” she warns. Industry bodies like APRA AMCOS predict generative AI could cannibalize up to 25% of the music industry’s income. “This is our opportunity to try and rewrite that history,” she adds. “We can pilot frameworks that actually feel fair from the start, rather than waiting for the damage to be done.”
Attribution black box: The whole system breaks down at the attribution puzzle, which the industry has failed to solve. The ambiguity is clear in volume-based streaming, where an artist can get a hundred thousand streams and earn only a thousand dollars. With AI, the math becomes even more impossible. “It begs the question: If you use an artist's prompt and download the output, how is that artist compensated?” she asks. “Do they get a micro-fee? Do they get 100%? It's just… no one really knows at the moment.”
Rejecting the soup: The ultimate vision, she believes, is a move away from monolithic AI systems toward a more artisanal ecosystem. “Right now, models like Suno are like a big bowl of soup, made of everything ever on the internet,” Ludwig says. “I imagine a world where people go to different platforms for their different ‘flavor’ or sonic essence.” She proposes starting with an “authentically Australian flavored model” as a pilot—one built on consent, where artists are paid upfront for their data, and the output is curated to have a distinct character. “We can show the world a way forward that can be ethical, where you choose a model based not just on its power, but on its values and what it offers creatively.”
A small circle fight: Ludwig rejects the retrospective, top-down model of big tech, where companies build massive models and deal with the legal fallout later. “I come from the perspective of, how can we work together from both the bottom up and the top down?” she says. “That is, going from a community level to the institutional level to a government level, and working together to solve this problem.” She also points to an 'education piece' that's missing, leaving many creators unaware of the implications of data scraping. The fight, she notes, is currently being waged by a "very small circle" of insiders.