Michael Smith, 54, of North Carolina, was sentenced to 18 months in prison after pleading guilty to a scheme that used artificial intelligence and bots to manipulate music-streaming services and steal royalty payments. The court also ordered him to forfeit $8,091,843.64 that he obtained from the scheme.
The U.S. Department of Justice said Smith is the first person to be criminally charged in the United States with AI-assisted streaming fraud. The case began with his indictment in 2024, but the scheme had continued for seven years since 2017 without being detected.
How did the scheme work?
Smith created thousands of fraudulent accounts on Amazon Music, Apple Music, Spotify, and YouTube Music, using fake email accounts and debit cards obtained fraudulently. He then ran software that caused bots to stream songs he claimed to own, while the Department of Justice confirmed that the songs were generated by artificial intelligence.
By the time of his arrest, he had created hundreds of thousands of AI-generated songs and distributed billions of automated streams among them. According to authorities, distributing the streams across a huge number of songs helped avoid drawing attention to a single song receiving an unusually high number of plays.
Numbers that surpassed global artists
Department of Justice data illustrates the scale of the activity: during April 2023, the fake songs received 80.9 million fraudulent streams, while Taylor Swift's entire catalog received approximately 9.3 million streams during the same month.
The scheme's impact is not limited to streaming platforms. Royalties are distributed through shared royalty pools, meaning that the fake streams take a portion of money that could otherwise reach real artists. The department says this reduced royalties across the ecosystem.
Why does this news matter?
The case reveals that combining automated content generation with account creation and automated operations can turn fraud from limited manipulation into a large-scale activity that is difficult to detect when the operations are distributed across a large number of files. The case also provides an early test of how law enforcement agencies will handle the use of artificial intelligence in financial crimes targeting digital platforms and revenue-sharing mechanisms.
The details of the detection tools used by the platforms remain undisclosed in the available material, so the case alone does not establish how capable the services are of preventing this type of attack in the future. But it shows that the risk concerns not only the creation of fake songs, but also the large-scale coordination of content, accounts, and automated plays.
The sentence and the dispute over it
Smith's lawyers requested probation instead of prison, citing his remorse and a job offer paying $5,000 per month in audio production, mixing, mastering, and recording. They also argued that Smith did not actually retain the full $8 million and that holding him solely responsible for the music industry's problems represented a model punishment.
The Department of Justice rejected that argument, saying Smith used increasingly deceptive methods and lied to police when confronted, and that a stronger sentence was necessary for deterrence. In addition to prison, the sentence required him to forfeit the funds that authorities attributed to the scheme.