Over the last year, I have experimented with AI music tools like Suno and Udio, where you can type a short prompt and get back a fully produced song. Within seconds the system generates lyrics, melody, instruments, and vocals for you, whether you want it to be your own or someone else’s. It is like having an entire recording studio on your laptop. But as exciting as this is, it also raises a number of questions about originality and ownership.
On the one hand, AI is making music production more accessible. You do not need expensive gear, years of training, or even the ability to sing. A student in Rotterdam can now create a radio quality track in the same afternoon they finish an assignment. This could unlock creativity and inspiration for millions who otherwise might never have produced music.
On the other hand, the industry is facing serious challenges. Streaming platforms like Spotify have already had to remove millions of AI tracks flooding their catalogues. Record labels are in talks to license their music catalogues for AI training, but the debate remains: who should be paid when a model learns from an artist’s voice or style? Cases like the viral Fake Drake track highlight how easily AI can blur the line between tribute and impersonation.
Currently, people are still able to notice the differences between most AI-generated songs and those produced by humans. AI music is often described as more meaningless and forgettable, though it is difficult to pinpoint exactly what makes it feel so empty. However, as AI continues to advance, the line of recognition is becoming increasingly thin. Though, I do wonder whether originality and emotional depth can truly ever be replicated.
I think that one possible improvement would be to require AI music tools to embed a subtle watermark or trademark within generated songs. Similar to how some artists use signature sounds or producer tags, this mark would make clear that the track was machine-generated. This would not prevent people from using AI music for inspiration or experimentation, but it would reduce the risk of someone passing off a fully AI-created track as their own original work. This way, listeners could better distinguish between human and machine creativity.
Personally, I found using AI to make music surprisingly fun, but also a little unsettling. When the song finished, I could not quite tell how much of it was mine. Did I really create it, or did I just write a clever prompt? If AI can generate a hit song from a simple text input, should we treat the human prompter as the true artist, or does creativity lose its meaning when machines do most of the work?
References:
Collins, K. C., & Manji, A. (2024, June 15–17). Humanizing AI Generated Music – Can Listeners Hear the Difference? In ResearchGate [Presented at the 156th Convention]. Audio Engineering Society, Madrid, Spanje. https://www.researchgate.net/publication/379671636_Humanizing_AI_Generated_Music_-_Can_Listeners_Hear_the_Difference
Spotify. (2025, 25 september). Spotify Strengthens AI Protections for Artists, Songwriters, and Producers — Spotify. Spotify Newsroom. https://newsroom.spotify.com/2025-09-25/spotify-strengthens-ai-protections/
Coscarelli, J. (2023, 19 april). An A.I. Hit of Fake ‘Drake’ and ‘The Weeknd’ Rattles the Music World. nytimes.com. http://nytimes.com/2023/04/19/arts/music/ai-drake-the-weeknd-fake.html
Great article! I think what you wrote is exactly what I have been thinking about recently. When AI is involved in artistic creation, how should we define originality, how should we define creativity, and how should we define artists? I recently followed a so-called AI graphic artist online. The pictures he posted were beautiful and avant-garde, and received a lot of likes, but there were a lot of arguments in the comments. Many people said that this was not artistic creation and called on people not to pay attention to such works; but there were also supporters who believed that this was also a legitimate means of artistic creation and we should encourage and support it.
I think that generative AI is undoubtedly a huge controversy in the field of artistic creation, which emphasizes originality. My view is that for works of art created by AI, whether it is music or pictures, as long as they can be loved by the audience, then they are at least successful in a sense. Let‘s continue to pay attention and see what will happen in the future.
While I would agree that AI tools promise great productivity enhancements in other fields, I think their use in music or other creative endeavours is a clear sign of not actually caring about that creative endeavour.
The creation of music is not only about the end product, but also about the process of actually making it. Its a process of trial and error that requires time, effort and a great deal of knowledge about the fundamental principles of music.
That is also why it is unlikely that you will create a good song with AI, much less a work of art. You do not have full control over the outcome, and most of the time you couldn’t even tell the AI what to improve about the song if you did. And if you did then you could probably do it yourself and would not need the AI.
Therefore, I think the decision to use AI for the creation of music makes it evident, that one does not care about the music at all, but only creates for an ulterior motive.
But I am really curious, what your motivations were to start making music, and why you chose to do so with AI tools instead of learning an instrument or downloading a DAW.
Interesting article Anna. The trade-off between the ease to create music, creativity and satisfaction is particularly one that I can recognize. Personally, I do not think that creating a song merely by AI would satisfy my desire to personal creativity or something that I would be proud of. However, I could understand that others would enjoy listening to certain tunes when the technology further improves and the music becomes even better. Therefore, the technology certainly has a huge potential.
Regarding the “watermark” I don’t feel the need to implement those. Why should there be a clear watermark to point out the difference between humanly made music and AI-generated music? As long as the listener enjoys it, music is music in my opinion. However, the AI-generated songs should not utilize a certain style or lyrics from human producers since that would profit from their talent and creativity.
All in all, really interesting. Curious where this technology could end!