How AI Music Caught Selena Gomez's Ear and Took Over Spotify

AI-generated music is no longer a novelty on the margins. It is pulling in tens of millions of YouTube subscribers, climbing global Spotify charts, and even surfacing in posts by pop star Selena Gomez. Inside the music industry, the warning is blunt: anyone treating songwriting and arranging as a mechanical task will struggle to survive. What is left for humans, many in the field argue, is the part that always mattered most: turning a personal story into melody.

The pressure is most visible at the working level. Generative music platforms led by Suno, along with a wave of new algorithmic tools, have reached a point where they threaten the livelihoods of human arrangers. One arranger in his late 20s said he is preparing to leave the field he built his career on. "Because of AI, I'm thinking about quitting music, which has been my way of making a living," he said. "Every new arranger is dealing with the same worry. For the price of a few cups of coffee, these tools can produce hundreds of songs. Who is going to pay many times that for a single track?"

The shift comes from a real jump in quality. Suno used to spit out low-grade audio with no commercial value. That changed in September 2025, when Suno V5 reached a level usable for demo tracks aimed at commercial release. For rookie artists who do not need top-tier audio fidelity, the tool can already stand in for a human arranger.Singing ability is no longer a barrier either. Composers who cannot sing do not need to book a vocalist. Suno can generate polished vocals with strong range and expression, and it can also turn a composer's own voice into an "AI persona" capable of pro-level delivery.
The old complaint that AI melodies sound flat and lifeless has lost much of its weight too. In 2024, Suno added a "Cover" function that builds on rough ideas a human writes first. A songwriter brings the melody, the lyrics, and the direction. Suno handles the arrangement and produces the finished sound.
The squeeze is now spreading to instrumentalists. On January 14, an algorithm-based instrument performance program called Tonalic launched, trained on a huge body of original performance data. It can imitate the small touches and nuance of a human player, narrowing the space session musicians have long owned.

Earlier this year, listeners were already crossing the line without thinking about it. AI-generated music was being streamed worldwide alongside human-made tracks with no clear distinction. As of January 23, two songs by AI artist Sienna Rose sat on Spotify's Viral 50 - Global chart, which combines social sharing and streaming data from Spotify users worldwide. Her signature track, "Into the Blue," peaked at No. 10 and was sitting at No. 28. Another song, "Safe With You," sat at No. 35 on the same chart. A third track, "Where Your Warmth Begins," was not on the global chart as of the 23rd but ranked No. 28 on the Viral 50 - USA chart, a snapshot of momentum in the American market.
Selena Gomez briefly pulled Sienna Rose into the spotlight when she used the artist's song "Where My Heart Unwinds" as background music for a Golden Globes-related post. After the AI-music backlash spread online, Gomez deleted the audio, which only drew more attention to it.
The first sign that AI music could attract a real mainstream audience came from the band The Velvet Sundown. The act racked up over 1 million plays on Spotify last year before it was revealed to be an AI artist, drawing a wave of fresh coverage.
YouTube has become the other major front. The biggest AI music channel, Masters Of Prophecy, produces hard-hitting electronic tracks entirely through AI and has pulled in more than 35 million subscribers. In South Korea, the scene looks different. There is no breakout domestic AI artist on the scale of overseas acts, and most listeners encounter AI music as background audio on YouTube. The Korean AI music channel SISO Wave has around 113,000 subscribers, while bgmMP3 has about 46,400.

The debate over whether any of this counts as "good" music is also moving in real time. On the global forum Reddit, a recent thread arguing that AI music is already better than most songs on streaming platforms drew more than 500 replies. The original poster pointed out that more than 86 percent of tracks on major streaming services have been streamed fewer than 1,000 times. "Just because a song is released into the world doesn't mean it's high quality," the user wrote.
One reply pushed back. "People look for music that excites or moves them, even if the sound quality isn't great. Music that feels familiar and fresh at the same time. AI can still only make the familiar part. Even if AI music sounds clean, I don't think it can make truly great music."
Several working professionals in the music industry said the answer is less about the technology and more about the person using it. "What matters in the end is who uses AI and how," they said. "These tools are efficient aids that help humans." Their point is that any AI-generated track with real popular appeal is not the machine's work alone. "Humans have to take the lead in editing. That's how quality music actually comes out," they added.
Industry figures also believe AI is unlikely to take over the human composer's space, at least for the next few years. As long as AI handles the writing of melody and lyrics on its own, the results will lack a distinct voice. Generative AI is built to produce the "statistically most likely" answer, which by design pulls every song toward the middle.
That gap is where the human role is being redrawn. The mechanical parts of music making, the parts that can be templated and repeated, are now squarely in play for AI. The part that still belongs to people is the one that was never about speed or volume in the first place: a writer with something specific to say, putting it into a melody only they could have written.
Minkyung Lee [email protected]