Are you listening to AI-generated music without even knowing it?.

How many hours of AI-generated music do you listen to per week? If I was to guess, “0 hours” would be your confident answer - maybe aside from the silly songs that accompany memes on Instagram.

We'd all like to bet on our skills to detect AI but when it comes to music the situation is far more complex. Producer, mixer and As I Lay Dying bassist Chris Clancy seems to believe you may be listening to more of it than you'd think…

A SONG ABOUT COFFEE

“I remember being in my living room with some of my best friends and one of them said, ‘[AI can] never catch the emotion of a performance.’ So I made a song about drinking coffee in the style of a 90s love ballad. It was such a great vocal performance. It was heart-breaking. I couldn't have sung it as well as AI made it. It was so emotionally provocative. It was a perfect performance and no one could believe it.

“The worst part is that while there are a lot of people who will always want authentic real music - the lines are not only already blurred, they're going to be so much more blurred in the coming years. There's just no way people are going to know.”

Unfortunately for appreciators of organic music, Chris is right. In April 2026 Deezer reported that 44% of all newly uploaded music to the platform is AI-generated, accounting for 75,000 tracks per day. Chances are high that you can spot this stuff from a mile away, but oftentimes you won't even know. Similar to CGI, you don't notice it's there if well-done. From cars to pylons to entire backdrops - assets can be added, changed, removed and the viewer will have no idea. To use an example a bit closer to home, we have autotune.

The classic ‘have/haven't they’ autotune debate has been prevalent since Dr Andy Hildebrand brought the technology to the table in 1997, popularised a year later with Cher's Believe. Ultimately it is the uncertainty on its use that allows the most worshipped of artists to silently get away with using it like a crutch.

Chris claims that playing live is the only thing AI cannot threaten to take away.

“People can't hear autotune on Justin Bieber and they'll argue with you. There are entire YouTube channels about his untuned live vocals - ‘listen how amazing he is, he doesn't even need it.’ He's auto-tuned within an inch of his life. But people can't hear it. So how are they going to hear AI?”

Moreover, many fans can be so devoted to the image of their favourite artists that they are willing to convince themselves that the singer(s) aren't using autotune so as to not shatter the world they have built for themselves. So while we associate AI music with the likes of the infamous and clearly artificial ‘We Are Charlie Kirk’, how many actually well-polished AI tracks are flying under the radar and out of our speakers? A lot, according to Clancy:

“I'm not going to name any names. But there are charting American bands using AI to generate tracks - not a few, a lot.

“I've never looked at the stats but I'm hearing bands people are going to know. People don't know [that it is AI]. People haven't noticed.”

Services such as Spotify have shown no interest in banning AI-generated music, only stepping in if copyright issues become present (if a song imitates another singer's voice, for example). In fact, in 2025, three AI-generated tracks topped their music charts. On top of this, a Deezer and Ipsos study collaboration discovered that 97% of people are unable to tell the difference between human-made and AI-made music.

You may be confident that no LLMs were involved in the creation of your favourite pop star's new record. But it's important to remember: if they did it right, you won't know they've done anything at all.

SO WHERE WILL THIS TAKE US?

Similar to AI's presence in every other industry, it's hard to know. But over the next decade there is a clear path to how it can ease the music workflow and produce tracks without the blood, sweat and tears currently shed to piece them together.

“Eventually fans will record the guitars and the bass, the drums. AI will help suggest parts - it'll be baked into the DAWs. AI might say, ‘this drum set sounds a bit busy, maybe you should try this.’ And if the guitars are played badly, it will just generate better performances. You'll say, ‘I want it to sound like Trivium,’ and it will generate in that style.”

It's a strange pill to swallow. But at the same time it's not like music recording and computer tools haven't already been working alongside one another for decades. So is AI simply the next step in that partnership?

Maybe, maybe not. “I frequently get people submitting: ‘here's my demo folder, it's 12 tracks of entirely AI-generated music’ and they want me to program the drums, learn and play the guitars, track the bass then release it as an album under their name. And at that point I'm like, that's too much for me.

“A lot of people on mixing forums outright refuse to work with anyone who's used AI. To me that's clinging to morals, but the world's just going to change. You go with it. If you're excluding all these people, it's just not going to work.”

“If there's a genuine performance in there and they've used it as a tool - that's fine. But if they just generate the track and say, ‘I've got a track’ - it's not doing anything.”

Ultimately it is a case where ‘only time will tell’. As a current example we can see that while recording straight to tape remains possible, it is an unnecessarily long-winded method for many. In ten years time, will we look at non-AI aided music through the same lens?

I suspect we may even see producers and mixers the same way as rear-view mirrors in cars. Racers used to have ‘spotters’ to look behind and see if anyone was overtaking and, if so, where from - until the rear view mirror was invented in 1910. Sure, you could have a man regularly looking behind to see if there is a car behind you, but it's a bit excessive in today's day and age, no?

Chris is glad he is a part of As I Lay Dying as it secures him a job if AI makes him redundant. But who knows - maybe there will be a revolutionary breakout band that decides to do it the old fashioned way. I suppose we will have to wait and see.