Anyone who spends a portion of their day scrolling on TikTok, as I do, has probably heard the song “I Run.” It’s the latest in a line of instant hits, its popularity ignited by a platform that thrives on 10-second sound bites. Every TikTok user knows what I mean: a snippet of a song goes viral, and suddenly it’s in the background of every video, permanently stuck in your head. This was my exact experience with “I Run.” It was catchy (honestly, too catchy), and one night, I couldn’t stop the endless loop of its chorus circulating through my brain. Upon mentioning this to my roommate, she replied, “Did you know that song is AI?”
This shocked me to my core. Not only was I unaware that the song was made using AI, but I never would have suspected it. Since this realization, a song I once tolerated (or maybe even liked) now fills me with anger and frustration. Have we really reached the point where artists use AI to replace human voices? And how was I completely unaware? I decided to dive into the creation of this song to discover how much of it is really AI-generated, as well as the implications of this phenomenon on human artistry and expression.
What initially raised suspicions of potential AI use was that the track features female vocals, with no credit given to any female artist. “I Run” was released by new British musician Harrison Walker, publicly known as HAVEN., and co-produced by a collaborator who goes by Waypoint. Listeners online speculated that the female vocals sounded similar to those of British R&B singer Jorja Smith, but she quickly clarified that she was not featured on the track. This is when the rumors started to swirl. Allegations surfaced that HAVEN. used AI to create a “deepfake” of Smith’s voice, meaning that her vocals were replicated with AI software. HAVEN. denied the claims, saying the female vocals were his own voice processed and filtered through many layers of audio editing. But that audio editing software? Waypoint himself confirmed that it was, in fact, a generative AI platform called Suno, which is currently facing copyright-related legal controversies.
Amid these AI concerns, along with concerns from Jorja Smith’s legal team about impersonation and copyright infringement, “I Run” was removed from streaming platforms. The song was also withheld from the Billboard charts, even though it was expected to perform well due to its success on TikTok. HAVEN. continued his attempts to re-upload the track, but it was removed each time.
The story doesn’t end here. Enter Kaitlin Aragon, a creator who posted a cover of herself singing the female vocals of “I Run.” In this video, she responded positively to commenters urging HAVEN. to officially add her voice to the track. So they did. The song “I Run” has since been re-uploaded to streaming platforms by HAVEN. and Aragon. Clearly, the addition of a real voice was an attempt to rescue his career following backlash over the AI use. But this controversy is only the beginning of a much larger problem surrounding the role of AI in art.
Primarily, I appreciate that streaming platforms and other music services seem to be taking AI use seriously. Many appear to agree that generative AI has no place in music, especially when it replicates the likeness of a specific artist, like Jorja Smith. But what if that changes? What if it becomes acceptable for producers to favor AI over real human vocals?
This is especially concerning given the rapid development of AI. While some listeners recognized that something was off about “I Run” and voiced their suspicions, I would guess that far more people did not notice an issue. I myself was one of them. It’s unsettling how undetectable AI is becoming, and it will only advance with time. If society begins to embrace AI-generated music, it will be nearly impossible for those who don’t support this practice to differentiate between the two.
Ultimately, the problem with AI-generated music is that it is an attempt to replace real human expression. Songs are stories. They are crafted from experiences such as heartbreak, friendship, love and joy. These are human emotions, ones that no robot could truly feel. But AI can replicate them, and it can do so almost instantly without ever requiring compensation. If society allows AI to creep into the music industry, there will be no room for people with authentic stories to share.
Needless to say, I am concerned about the future of AI. But for now, I think there’s still an opportunity to curb this dilemma before it becomes unmanageable. In my opinion, that means stricter regulations on AI, not just in music, but in all forms of art. People should continue to call out content they suspect to be AI. It’s important to hold artists accountable to ensure artistic integrity and preserve the true purpose of music. I think society will eventually find the place where AI belongs, but that place is definitely not within the music industry.