Artificial intelligence is already reinventing the way we write, draw, and film. Now, it is venturing into a new highly symbolic territory: music. If OpenAI is indeed developing an AI capable of composing on demand, what future is left for musicians, composers, and sound artists?
OpenAI is developing a musical AI capable of creating pieces from text or audio
A persistent rumor, relayed by The Information and TechCrunch, suggests that OpenAI is working on an artificial intelligence capable of composing music from a simple text or audio prompt. Imagine this: you type, “a melancholic folk ballad with vocal harmonies,” and the AI delivers a complete piece.
This does not come out of nowhere. OpenAI has already experimented with music through MuseNet and Jukebox, but this would be a more advanced, accessible technology, likely integrated into Sora, its video generator. The idea? Automate the entire audiovisual creation process, including music. To train its future AI, OpenAI would be collaborating with students from a major American music school. In other words, the die is cast.
If this AI becomes a reality, what risks do musicians face?
Let’s be honest: the arrival of this technology could shake many foundations. A composer who takes 15 days to create a piece versus an AI that does it in 30 seconds? It’s challenging to compete on the productivity front. And what about music generated on-the-fly for ads, games, and social videos?
More seriously, some raise concerns regarding the training data. If the AI learns by ingesting existing pieces without authorization, this raises significant questions about copyright and ethics. Then there’s the symbolic value: can a melody generated by a machine equal the intent of an artist, their flaws, their intuitive choices?
What if this same AI opens unexpected opportunities for creators?
But not everything is so bleak. I once had to use an AI to test a melody for a project video. The result? A good base… which I quickly remixed, arranged, and twisted. The tool does not replace; it catalyzes.
This technology could enable artists to prototype, explore new styles, and save time. Moreover, the more prevalent AI becomes, the more the audience will seek authenticity, sensitivity, and the human touch. Musicians could then leverage this distinguishing factor as a form of poetic resistance.
New professions might emerge: musical prompt designer, AI sound curator, or even “augmented composer.” Labels might offer hybrid albums: AI + human, fully transparent.
Artists still have a chance, provided they adapt and embrace their humanity
So, are artists doomed? I don’t believe so. But their profession is changing, irreversibly. Those who can harness these tools to enhance their creativity will have an edge. Those who focus on the human value of their music will, too.
We are entering an era where AI can compose. But it cannot feel, doubt, remember, or dream. That might be the key.



