The Machines Are Singing Now
Setu Garg, Director of Marketing | Published on 10/1/2026
I'll be honest. When I signed up to write this article a few weeks ago, I wasn't entirely sure how I felt about it. AI-generated music felt like something that happened to the music industry, not something I needed to engage with. Then I listened to the Vergecast episode "The problem with Suno and AI music" from July, where Terrence O'Brien joined David Pierce to dig into where Suno came from, how it's being used, and whether it actually changes anything. I came out of it genuinely unsettled. Not in a bad way, but in the way that happens when a topic you'd been comfortable ignoring suddenly refuses to stay in the corner.
So here's where I landed. And I want to be upfront: this isn't a neutral take.
What Music Actually Does for Me
A big part of why I love music is the story behind it, which has nothing to do with the sound itself. And it's why I'm an audiophile in the first place.
When I'm listening to a track I love, my mind wanders into the creative process. How did the artist arrive at that arrangement? What was going on in their life when they wrote this? How did they decide that this instrument, that this tempo, was the one? The music becomes a window into a human mind working at the height of its creativity. That's what moves me. This is what I'm chasing. The creativity that comes out of the complex arrangement that Hans Zimmer and A R Rahman do in their tracks makes them my all time favorite composers.
PS: Both these legends are collaborating for the soundtrack of the upcoming Indian mega production “Ramayana”. Totally worth checking out the soundtrack as it comes out this year!
AI, by definition, can't give me that. It isn't drawing from a life lived or an emotion felt — it's drawing from a vast catalogue of everything it has been trained on. No matter how convincing it sounds, I know what's behind it. And for me, that changes everything.
The Moment It Became Real
A few weeks ago I came across an electronic track on YouTube Reels. Catchy, well-produced — exactly the kind of thing I'd normally go chase down on Spotify. So I did. And right there on the listing was an AI label.
I was immediately disappointed. Not because the music was bad. But because the thing I was about to do next - look up every artist involved in the track, explore their other work, find albums I hadn't heard, and add them to my library, suddenly had nowhere to go. That whole thread just... stopped.
That's my ritual when I discover something I like. I follow the humans behind it. I build a relationship with their catalog. I take pride in what I've assembled over the years and what it says about the artists I've chosen to follow. With an AI track, that entire process collapses. There's no artist to discover. No body of work to explore. No story to follow.
What Is Out There?
For those curious about the tools driving this shift, here's a quick lay of the land.
Suno is the most accessible entry point — type a sentence, get a finished song, vocals and all. The latest version now includes DAW-style editing tools signalling this is no longer just a novelty. (DAW stands for Digital Audio Workstation, the software professionals like Logic Pro or Pro Tools use to record and mix music.) The legal situation around training data is still unsettled, and purely AI-generated works can't currently be copyrighted under US law.
Flow Music is Google's more musician-focused answer — a full workspace for generating, editing, and producing tracks. What's interesting about Google's approach is its framing around collaboration rather than replacement. Grammy-winning artist Wyclef Jean used the underlying model as a creative tool while making his song "Back From Abu Dhabi," treating AI outputs as raw material to shape rather than finished product to ship. He put it well: "There's one thing you have over the AI: a soul. And there's one thing AI has over you: infinite information." Google also showcased the tool with legendary Indian composer Shankar Mahadevan through their India blog — worth a watch.
I respect this approach more than fully generated music. But I'll admit I'm still not sure where I land on it.
What Are the Streaming Platforms Doing About It?
The scale of what's happening is worth pausing on. Deezer reported that AI-generated tracks peaked at 90,000 per day in June 2026 with over 50% of total daily uploads on their platform. Apple Music disclosed that over a third of its new uploads are now fully AI-generated. Spotify deleted more than 75 million AI spam tracks in a single year-long crackdown.
Yet listeners aren't actually choosing to hear it. AI tracks account for just 1 to 3 percent of actual streams on Deezer, and under half a percent on Apple Music. The flood is real. The demand isn't. Most of that "listening" turns out to be bots padding play counts, with 85% of AI streams on Deezer flagged as fraudulent and demonetized.
- TIDAL is the most aggressive. Starting July 15, 2026, TIDAL labels wholly AI-generated music and excludes it from royalty attribution — the first major platform to move from disclosure to demonetization. I appreciate that.
- Qobuz — the platform many of us lean on for hi-res — has taken a human-first stance, building its own detection system and keeping editorial picks entirely human-selected.
- Spotify recently announced "AI Persona" badges for AI-generated artists, with those tracks excluded from algorithmic recommendations by default.
- Apple Music relies on uploader disclosure — which, as you might guess, is the weakest of the approaches.
The honest problem: a listener who wants to avoid AI music still has no foolproof way to do so, because disclosure is largely voluntary. The labels are a start. But this needs to go further.
An Ethical Problem, Not Just an Aesthetic One
My resistance to AI music isn't just about the listening experience. It's about where the money goes.
Streaming platforms already don't pay artists what they deserve — that's a well-documented and ongoing injustice. AI music is going to make that worse. If a generated track can fill catalog space and attract passive streams without a human artist ever seeing a royalty, the economics of music creation get even harder for the people who actually make it.
This is part of why I buy physical media. It's a small act, but it's a direct one — money going to artists in a way that streaming rarely manages. And it's one of the reasons I genuinely appreciate platforms like TIDAL and Qobuz drawing a line.
Quality over quantity has always been my philosophy when it comes to music. A carefully assembled library of music I've sought out, from artists I've followed, means something to me.
What This Means for the Foundation's Community
We're a community built around the love of music and the pursuit of better sound. AI music doesn't change that mission. If anything, it sharpens it.
A few things worth keeping in mind: The flood of AI-generated filler may actually push more serious listeners toward physical formats and hi-res platforms with stronger curation, which is good news for the vinyl renaissance which John Hughes covered in the June's Muse Intro.
Transparency matters. Support the platforms and artists who are upfront about AI's role in what you're hearing. Our ears and our dollars deserve to go where they mean something.
And if you haven't listened to that Vergecast episode yet, give it a go! It won't tell you what to think about AI music. But it'll make sure you're thinking about it.