Imagine you're a global superstar who has spent years perfecting your songs, only to discover that those exact tracks may have been secretly used to train an AI now capable of producing music in your style. It is precisely this anger that has driven two of the biggest names in the music world to the barricades in recent days: SZA and Doja Cat.
SZA's scathing takedown
Over the weekend, SZA posted an unusually direct message on Instagram. She alleged that producer Diplo is an investor in an AI music company and helped train its models using the works of the "best and brightest Black songwriters and producers." She claimed that hundreds of her own songs were used for this training.
Her choice of words left no room for interpretation: she labeled musicians who support AI music generators as proponents of "degenerate shit" and called them "disgusting." Her concluding sentence was particularly poignant, wishing those responsible "get the life they deserve"—coupled with the clear statement that nothing could ever justify this practice in her eyes.
Doja Cat: "Fuck AI for real"
Just a few days later, Doja Cat joined the criticism. The catalyst: fans had mistaken AI songs circulating online for her own music. Her reaction was just as blunt as SZA's—short, clear, and leaving no doubt about how she feels regarding AI-generated music.
The other side's response
Faced with these accusations, a spokesperson for the company involved did not respond directly, instead pointing to a LinkedIn post by their internal product lead. It stated that the training data does not contain artist names, the system cannot reproduce exact content from the training material, and they are working increasingly hard to better detect identity misuse.
Ironically, this controversy lands in the middle of a phase where the company in question just introduced a new funding program for independent artists—and was attempting to publicly position itself as being particularly artist-friendly.
Why this dispute is bigger than it sounds
SZA and Doja Cat are by no means the first stars to speak out critically against AI music; Paul McCartney and other well-known names have previously expressed concerns about the future of intellectual property in the music industry. However, when two such influential, young artists take a public stand within just a few days, it shows that the issue of AI music is no longer just a niche debate among lawyers and tech enthusiasts—it is directly and emotionally affecting the industry's biggest stars.
What this means for you as a listener
This story highlights just how deep the divide between established music stars and the world of AI music truly is. While you as a listener might enjoy creative AI experiments, the anger from SZA and Doja Cat shows that behind every song used for AI training, there are real people with real feelings and real labor. For us at AI.OnAir, it is clear: this exact tension between creative innovation and the protection of human art will continue to define the music world in the coming years.
Conclusion: What this means for AI.OnAir listeners
The public outburst of two global superstars shows that AI music has long since arrived at the heart of pop culture—not just as a trend, but as a tangible point of contention. We at AI.OnAir remain eager to see how this debate evolves and will provide updates as soon as there is news in this dispute.
Get your music on the radio
Are you ready to present your songs to a wide audience? Submit your tracks via the AI.OnAir Control Center, benefit from heavy rotation, and build your own premium artist page.
Jetzt Song einreichenFrequently Asked Questions (FAQ)
The stars criticize that their music was used without permission to train AI models. They view this as a violation of intellectual property and find the use of their style disrespectful toward their creative work.
Doja Cat reacted angrily after fans mistook AI songs circulating on the internet for her own new music. This highlights the issue of identity confusion caused by AI technology.
The companies involved often claim that their training data does not contain artist names and that the system cannot reproduce exact content. Furthermore, they point to measures taken to detect identity misuse.