Eight million dollars in stolen royalties, hundreds of thousands of fake songs, and billions of bot-driven streams. That's what Michael Smith, 54, of North Carolina built over seven years before pleading guilty this week to wire fraud and money laundering conspiracy.
Smith's operation was simple in concept and massive in execution. Between 2017 and 2024, he used AI music generation tools to create a catalog of hundreds of thousands of tracks, then uploaded them to Spotify, Apple Music, Amazon Music, and YouTube Music. To generate plays, he ran up to 10,000 bot accounts simultaneously, purchasing bulk email addresses and outsourcing the account registration to hired labor. Automated software directed the bots to stream his tracks continuously, while VPN routing disguised the traffic as legitimate listeners spread across different locations.
The $8 million in fraudulent royalty payments came directly from pools meant for real artists and songwriters. Every dollar Smith collected was a dollar taken from someone who actually made music.
How He Stayed Under the Radar for Seven Years
Smith's evasion tactics were more sophisticated than the typical bot farm. Rather than concentrating streams on a few tracks (which platforms can easily flag), he spread plays across his massive catalog of AI-generated songs. The VPN routing added another layer, making the bot traffic look geographically distributed. According to the case details, a co-conspirator and the CEO of an AI music company helped him acquire the generated track catalog at scale.
The streaming platforms eventually caught on, but not before Smith had collected millions. U.S. Attorney Jay Clayton noted that while the songs and listeners were fake, the money was very real.
What This Means for AI Music Tools
Smith faces up to five years in prison. But the case raises a bigger question for anyone building or using AI music generation tools: the streaming platforms still don't have reliable ways to distinguish AI-generated content from human-made music at upload time.
This wasn't a case of someone using AI to assist with songwriting or production. It was pure manufacturing - content created solely to extract money from royalty systems that assume each stream represents a real person choosing to listen. As AI music tools get cheaper and better, the incentive for copycat schemes only grows.
Spotify and others have tightened their fraud detection since Smith's operation was uncovered, but the fundamental vulnerability remains. Royalty systems were designed for a world where making music required effort. AI removed that constraint, and the platforms are still catching up.