On Oct. 14, 2025, music artist Breaking Rust released their song “Livin’ on Borrowed Time.” This song quickly rose to popularity, gaining over 22 million streams on Spotify and over 21 million on YouTube.

Except, Breaking Rust isn’t actually a real music artist, and “Livin’ on Borrowed Time” isn’t a real song.

They are both AI-generated.

This song is just one of several chart-toppers made by AI.

And while some may argue this song only gained popularity once people knew it was AI generated, that itself is also problematic. The issue is that people have a dangerous mindset — they think of AI music as a silly, harmless little thing.

This isn’t just some fun gimmick though; it signals a bigger issue emerging within the music scene.

AI is detrimental to the music industry and is bringing about sweeping changes that will affect anyone who consumes or produces music.

The lack of firm legal guardrails around this will fundamentally change an already struggling industry with ease, like a warm knife through soft butter.

And I don’t mean the famous artists or streaming services, I mean all musicians in general — most of whom are unable to make a living wage.

But first, to understand the detrimental effects, it is crucial to understand how AI music works.

AI music generators, like most other AI models, need training. They are fed hundreds of thousands of tracks, often without consent from the artists, and they break them down into patterns, such as chord progressions, tempo changes, and note arrangements.

Then, they use deep learning to recognize the significance of these patterns — what notes sound good together, how different instruments blend, and what emotions they evoke.

Finally, the AI analyzes the prompt and uses its stolen knowledge to generate a song that matches it, turning a creative process that has historically taken talented musicians months — sometimes years — to accomplish into a minutes-long affair.

But the truth is, the music isn’t as “AI-generated” as it claims to be. While the model seems like a composer, what it is really doing is stealing real musicians’ work and stitching it together. 

Take Breaking Rust, for example. Real musical artists Blanco Brown and Bryan Elijah Smith have claimed that another song by this AI persona is based on them, using their vocal style and artistic identity without consent or credit.

This reveals another problem: copyright. Courts are still divided as to whether the use of human musicians’ work for training AIs falls under “fair use” or “copyright infringement,” leaving art unprotected and exposed until a firm conclusion is reached.

Personally, I believe this is an example of copyright infringement. Sure, musicians take inspiration from each other all the time, but it doesn’t work the same way with AI.

Most humans don’t have the capability to deeply analyze patterns across hundreds of thousands of songs and recreate someone’s entire manner.

Real musicians build off each other; AI just copies.

The existence of AI “songs” directly disrespects the hard work that real artists put into making their music. When their hours of work are reduced into tokens to be played for the whole world, it signals a huge flaw in the system.

But this doesn’t only affect music already made; it warps the music of the future.

Now that AI can produce tracks in minutes, the artistry of human musicians — who already have a difficult time finding work — may be snubbed in favor of profit.

In creating music, there are a lot of people to compensate and manage: the songwriter, arranger, musicians, producer, recording engineers, and mixing engineers. Already, companies are turning to AI to generate music for their products, like video game soundtracks or commercial jingles.

With AI there’s only one entity on the payroll, and it only costs about $10 a month to use tools like Suno AI. To someone concerned solely with efficiency and profit, AI is the obvious choice.

Now, all these roles — composers, performers, audio engineers — will become more and more obsolete as AI uses their predecessors’ work to do their jobs faster.

To avoid this, I think that copyright protection needs to be applied for all songs used in training, and musicians need to consent to their music being used, or at least be compensated to some extent.

Even then, it will still cause issues.

Even then, real musicians’ jobs will be taken.

Because at its heart, this is not a problem of laws; it is a problem of morality and society’s interpretation of art.

Right now, the quick and unquestioning acceptance of AI music is frightful. People view it as a gimmick rather than something that can seriously affect their listening consumption and the livelihood of artists.

Before we know it, AI music will creep up on us, and one day, without realizing it, we may be listening to music and hear an AI “song.” Already, Spotify is recommending them to people.

But if you do encounter one, don’t think of it as just another song. If you use a music generator, don’t think of it as just a fun toy to play with.

Realize what this means for us as a society and for the people who have dedicated their whole lives to music — people whose passions and careers are becoming more and more precarious — and whose music you are actually listening to.

Remember and support the real musicians who bring life to the music industry.