First AI streaming fraud guilty
A North Carolina man, Michael Smith, pleaded guilty in the first-ever federal case for using AI to create hundreds of thousands of fake songs and inflate streaming plays — the scheme reportedly netted him over $8 million in royalties. ( ). Smith has agreed to return the illicit earnings and awaits sentencing, a legal precedent that could reshape how platforms and labels pursue AI-driven abuse. ( )
Smith agreed in his plea deal to forfeit $8,091,843.64 — the exact dollar amount listed in the court materials as the illicit proceeds to be returned. (thefederalnewswire.com)) Under the agreement Smith, 54, of Cornelius, North Carolina, admitted conduct amounting to one count of conspiracy to commit wire fraud, a charge that carries a maximum five-year prison term, and his sentencing is scheduled for July 29, 2026 before U.S. District Judge John G. Koeltl. (justice.gov)) Federal filings say Smith created hundreds of thousands of AI-generated tracks and used automated “bot” accounts to stream those songs billions of times across Spotify, Apple Music, Amazon Music and YouTube Music. (justice.gov)) Prosecutors’ estimates put the operation’s peak at roughly 661,440 fraudulent streams per day and indicate Smith deliberately spread plays across thousands of songs to evade platform fraud-detection. (musicbusinessworldwide.com)) The Mechanical Licensing Collective flagged irregular patterns and withheld payments in early 2023, a step that helped trigger the investigation that led to Smith’s September 2024 indictment. (musicbusinessworldwide.com)) Court records and reporting show hundreds of the tracks used in the scheme list Alex Mitchell, CEO of AI-music company Boomy, as a credited co-writer; Mitchell has not been charged in the case. (billboard.com))
