AI Remixes Are Stealing Record Sales From Original Artists

The California reggae band Stick Figure had a surprising problem: their song "Angels Above Me," released six years ago, recently shot to number one on iTunes in six countries, including the UK and Canada. But the band didn't make money from the surge. The chart-topping versions weren't the original track—they were AI-generated remixes that belonged to someone else entirely.

Four different unauthorized remixes of the song have gone viral. One version got 1.8 million YouTube plays in just five days. These remixes use artificial intelligence to change the speed, pitch, and arrangement of the original song, but keep enough of the melody and vocal to make it recognizable as Stick Figure's work. The band sees no payment from any of these plays.

How Streaming Platforms Respond

The remixes have spread across multiple platforms—Spotify, YouTube, and others—making it hard for the band's label, Ineffable Records, to take them down. Spotify did remove the unauthorized remixes when the label asked. YouTube and other services were less cooperative. When one remix creator was contacted, they claimed their version was a legitimate "cover" and offered to split the money. The band rejected this, saying a remix created by AI is different from a cover and should not have been released without permission and payment.

This problem has happened before. In 2022, musician Steve Lacy's song "Bad Habit" was flooded with unauthorized sped-up remixes on TikTok. The difference now is speed and ease. Old-fashioned remixes required someone with technical skill to manually edit audio files. AI remix tools let anyone generate dozens of variations with almost no expertise.

The Legal System Is Playing Catch-Up

The music industry has started fighting back. In December 2024, Universal Music Group sued companies that make AI music generators. Earlier, in June 2024, the Recording Industry Association of America (RIAA) sued an AI music service called Udio, claiming it can create songs that sound like famous copyrighted recordings when you ask it the right way.

The problem is that copyright law and enforcement tools were built for a different world. YouTube's system for catching copyright violations, called Content ID, is designed to detect exact copies of songs. It can't catch AI remixes, which sound new even though they borrowed from the original. The Copyright Office is studying how to handle AI and copyright, but rules haven't caught up with the technology yet.

The core issue is how AI music tools work. During training, these tools learn from millions of existing songs—including copyrighted ones. Then when someone uses the tool, it generates new audio that captures patterns from those copyrighted songs without copying them directly. Unlike sampling, where you can hear the original recording built into the new song, AI remixes create entirely new audio files that still manage to borrow the melody and structure.

The Enforcement Problem

For artists and record labels, the job has become relentless. Instead of taking down a few unauthorized versions here and there, they now have to hunt thousands of remixes across dozens of platforms. This creates an unfair situation: it takes minutes and almost no money to upload an AI remix to multiple sites, but it takes weeks and skilled staff for the record label to find and remove them all.

During that gap—sometimes days or weeks—the remix creator is making money from ads and streaming payments. By the time the track gets taken down, the damage is done.

I've watched technology shake up music before, and this feels familiar. When Napster let people share music freely in the 1990s, file-sharing moved faster than the law could handle. The same thing happened when YouTube launched and again when TikTok exploded. Each time, the industry eventually adapted—through new rules, legal decisions, or technology. But the adjustment always hurt artists first.

The underlying issue here goes beyond Stick Figure's case. If AI tools keep flooding streaming services with cheap remixes, listeners might play those instead of the real song. The original artist loses both money and attention. Multiply this across thousands of artists, and you start to erode the economic system that actually pays musicians to make new music in the first place.