A North Carolina musician has been sentenced to 18 months in federal prison for using AI-generated songs and automated listening accounts to collect more than $8 million in fraudulent streaming royalties. His operation generated billions of fake plays across major music platforms between 2017 and 2024, according to the US Department of Justice.
Michael Smith, 54, of Cornelius, North Carolina, received the sentence on October 6, 2026, after pleading guilty in March to conspiracy to commit wire fraud. By spreading automated plays across hundreds of thousands of AI-generated songs, Smith made suspicious listening patterns harder to detect while collecting royalties at the expense of legitimate rights holders.
How AI-generated songs fueled the fraud
Smith targeted services including Spotify, Apple Music, Amazon Music and YouTube Music. He created fraudulent subscriber accounts and used software to repeatedly stream music he controlled, generating royalty payments without genuine listeners.
At times, Smith operated as many as 10,000 bot accounts simultaneously, according to the Justice Department’s sentencing announcement. He used fake email addresses and fraudulently obtained debit cards to establish the accounts.
The scheme illustrates the difficulty of distinguishing genuine users from automated activity, a challenge other platforms are addressing through measures such as Facebook’s video-selfie identity verification.
Smith also relied on virtual computers, automated playback software and virtual private networks (VPNs) to run accounts and conceal activity, according to the original federal charges.
Generating enough music was essential to avoiding detection. Concentrating millions of fraudulent plays on a few songs could have triggered anti-fraud systems, so Smith distributed the activity across thousands of tracks.
He worked with an AI music supplier to build a catalog containing hundreds of thousands of songs. In emails cited by prosecutors, Smith acknowledged needing large quantities of music to circumvent streaming platforms’ fraud controls.
In April 2023, Smith’s accounts generated 80.9 million fraudulent streams through YouTube Music family plans. Taylor Swift’s entire catalog received 9.3 million streams in the same subscription category that month.
U.S. District Judge John G. Koeltl also imposed two years of supervised release and ordered Smith to forfeit $8,091,843.64.
Streaming fraud puts platform safeguards to the test
Smith’s March 19 guilty plea concerned fraudulent listening activity and royalty payments, not the creation of AI-generated music itself. Manipulated streams can divert royalties from musicians and songwriters whose recordings attract genuine listeners.
In July 2026, Deezer reported averaging approximately 90,000 fully AI-generated track uploads per day in June. Such tracks accounted for more than half of new music uploads on peak days.
Deezer also identified up to 85% of streams involving fully AI-generated tracks as fraudulent in 2025. Those findings apply to Deezer, not the wider streaming industry.
Other digital platforms are addressing different risks involving synthetic content. YouTube has clarified monetization restrictions for repetitive AI-generated videos, while Google’s support for European AI-content labeling standards reflects growing attention to identifying synthetic material.
For music-streaming services, identifying AI-generated songs alone cannot establish fraud. Investigators must also examine coordinated accounts, abnormal listening patterns and suspicious royalty activity.
Smith collected more than $8 million by distributing fraudulent streams across thousands of tracks. His sentence highlights the financial consequences of manipulated engagement and the need to identify suspicious activity before royalty payments are made.
Read more: Questions about ownership and commercial use of AI-generated audio are also reaching the courts, with a Japanese ruling on AI voice cloning recognizing legal protection for a performer’s voice under certain circumstances.