The future of country music is here, and it’s AI

When songwriter Patrick Irwin moved to Nashville last year, he was entering a lottery. Each day hundreds of sessions take place where writers create a song demo to pitch to a publisher. Publishers then share those songs with labels and managers, who may share those songs with the artists. Even if a major country star records (“cuts”) the song, it still takes a stroke of luck for that song to become a No. 1 hit.

The odds of winning are extremely low. Recently, Irwin was in a room where his co-writers Sam Fink and Duane Deerweater tried something new. Instead of booking studio time or calling a “track guy” to produce a demo, one co-writer opened Suno, an AI music platform, uploaded a voice memo with just guitar and vocals, and typed in a prompt: “traditional country, male vocal, folk country, story telling, 90s country, rhythmic.” 30 seconds later he had two fully produced demos complete with drums, electric guitars, bass, and backing harmonies. There were no studio musicians, no invoices.

“You tell it the genre and it totally does the whole thing, it’s insane,” says Irwin. He was as astonished as he was disturbed. This was not the Nashville, a city with a storied 200-year history of producing much of America’s greatest music, that he had imagined.

Irwin isn’t alone in this feeling. In the background, AI is taking over the city. At the start of 2024, few professionals had even tried these tools, but in the past six months, songwriters and producers have embraced them to work faster and cheaper — and for some, more resourcefully. No label, no major publisher, nor Suno would give comment for this story. But after speaking with musicians, writers, and a dozen anonymous insiders, it’s clear that Nashville has become an AI town.

Songwriter Trannie Anderson — who’s written for Lainey Wilson, Dan + Shay, and Reba McEntire — says the tech is ubiquitous. Though she doesn’t use it herself, she sees wide use “from entry-level songwriters to the top dogs.” She isn’t exaggerating; multiple sources have told me that even stars like Dustin Lynch and Jelly Roll are being sent pitches with their voices artificially generated into demos, something AI voice transfer makes possible. Lynch’s manager Brad Belanger confirmed this, adding, “What a world we’re moving into.” Jelly Roll’s representatives declined to comment on the record.

The headlines might be focused on “Walk My Walk” by Breaking Rust, an AI song that recently topped the Billboard Country Digital Song Sales chart, but that story is largely smoke and mirrors. That chart is a holdover from the pre-streaming era and easily gamed. The real revolution isn’t on the charts yet — it’s happening in the writing rooms.

Nashville is known as a “10-year town,” reflecting the amount of time you have to grind it out before getting a hit. Even for established songwriters, it takes a long while between the original songwriting session and making it to country radio. “Two years is the fastest I’ve seen,” says Jon Sherwood, a UMPG writer who penned the Bailey Zimmerman and Luke Combs hit “Backup Plan.” Sherwood still writes the traditional way, sans AI, but he notes the speed is changing the game. In the past, writers would pen a song, then pay a “track guy” $500 to $1,000 to record a professional demo to pitch to stars.

Suno is creating efficiencies in the Nashville song assembly line. Songwriters still write the lyrics and melody because Suno’s generated lyrics are cloyingly cliche (though many of Nashville-based sources agreed that country radio hits can be too). They instead use the AI to handle the demo production. They record a voice memo, upload it, and use the “remix” feature to turn it into bro-country, alt-country, or “hick-hop” in seconds. Maggie Reaves, a rising songwriter signed to publishers Dream 3 and Kobalt, recently had an assignment for a major artist with a one-day turnaround, so she wrote the song and “threw it in Suno.” Her publisher told her, “This is perfect. This is going straight to her.”

Reaves writes an average of 200 songs a year. Demoing her songs can be prohibitively expensive. She used to save up money to demo songs for $500 each — demoing all her songs could hypothetically cost tens of thousands of dollars each year. Now, she pays $96 a year for near-infinite attempts: “I immediately saw this could replace that.”

She’s not alone. Publishers are even running back catalogs through Suno to find new angles on forgotten tracks.

The sound isn’t perfect and can resemble an over-compressed, “dated” MP3: slightly lo-fi, not very dynamic, low sample rate and bit depth. But the uncanny valley of Suno songs is undeniably the voice, presenting with a slight grainy quality to the vocal that is overly pitch-corrected; the inflections are borderline Cylon. Yet Reaves says 70 percent of the output is solid enough to play in the car (traffic noise is enough to mask low-quality sound) and gives a clear picture of the finished song in order to pitch to an artist. These are demos, after all.

Suno isn’t just useful as a demoing tool; it also helps producers rapidly brainstorm different creative approaches to a song or musical passage. Independent songwriter Kalen Nash usually produces songs the old way, in his studio, track by track, but recently has adopted Suno for creative inspiration; he calls it a “band in your pocket.” He’s used it to turn diary entries into full songs. Jacob Durrett, a Big Loud producer, uses it to find alternative versions and “vibes” for songs. He can put in a “half-cooked” idea for inspiration: Just a guitar idea scratch track is enough for Suno to output multiple melody and full song ideas in any genre imaginable.  “I’m in awe of it sometimes, how good it can be, you know?” He says that Suno is giving him “a productivity boost more than a creative boost.” As a skilled musician, he’s equally capable of trying a song out in any style — it just takes longer. His hope is that AI will take over the tedious parts of his job, like renaming files and preparing them to mix, so he can focus on the creative part.

Music publisher Eric Olson, who encourages writers to use the tool, calls it an “unlimited co-writer in the room.” He finds it useful for coming up with samples without the headache of clearances, or concerns that someone else has sampled the same part. For him, it’s about buying back time. “If I can give Suno the last 20 percent and spend more time with my kids, that’s huge,” he says.

It’s not all biscuits and gravy, though. Most people in Nashville wear multiple hats: songwriter, “track guy,” studio musician, touring guitarist. “There used to be a whole world where musicians were making six figures only playing demo scale,” says Ian Fitchuk, the Grammy-winning producer of Kacey Musgraves’ Golden Hour. Fitchuk has steered clear of Suno and worries about the musicians losing that income. Trannie Anderson calls it “the final nail in the coffin” for the demo studio system. If the “farm team” of demo players disappears, the industry might face a talent crisis down the road.

There are also legal and ethical headaches: “If Suno spits out a lead line an artist uses, what’s the protocol?” asks Reaves. The copyright office doesn’t protect AI work, which makes the ownership of these hybrid songs messy; plus, the AI was trained on existing music. “AI learns from my songs, my friends’ songs … We aren’t being compensated,” says Anderson. And there’s the “ick” factor. Durrett hates when the AI outputs a voice that sounds exactly like a friend of his. (He claims it’s happened many times.)

Despite the concerns, Suno just secured $250 million in funding, is making $200 million in annual revenue, and adoption is accelerating. Wait around two years and you’re going to hear songs made with the help of Suno all over country radio.

The question is how they will sound. Anderson hears that “there’s an element missing,” she says. “Humanity and a soul … The Holy Spirit doesn’t live in AI.”