Loading page…
Prosecutors argued that automated listening drained royalties from real artists. Smith’s lawyers sought probation, saying prison would delay repayment.
Listen to this story
The operation used hundreds of thousands of AI-generated tracks and billions of automated plays, spreading activity across a vast catalog to evade fraud detection rather than concentrating streams on a hit. For April 2023, the DOJ’s narrow YouTube Music family-plan comparison put Smith’s bots at 80.9 million streams versus 9.3 million for Taylor Swift’s entire catalog; it does not compare total listening across platforms. The case ended with an 18-month sentence, more than $8 million in forfeiture and two years of supervised release.
Smith began the operation in 2017, using fake email addresses and fraudulently obtained debit cards to open thousands of accounts across streaming services.
In an October 2017 email, Smith estimated 661,440 daily streams and more than $1.2 million in annual earnings at half a cent per stream.
A 2018 email described spreading plays across many songs to work around platforms’ anti-fraud policies.
Michael Smith’s automated music operation has ended with an 18-month prison sentence and an order to forfeit $8,091,843.64. The 54-year-old North Carolina man pleaded guilty after using AI-generated songs and bots to collect streaming royalties. Prosecutors say the scheme took money from the shared royalty pools that pay genuine artists.
Smith must also serve two years of supervised release. He was indicted in September 2024 and pleaded guilty in March. The Justice Department describes him as the first American criminally charged with AI-assisted streaming fraud.
According to the DOJ, Smith began in 2017, using fake email addresses and fraudulently obtained debit cards to open thousands of accounts. The services included Amazon Music, Apple Music, Spotify and YouTube Music. Software then made the accounts repeatedly play songs he claimed to own.
By his arrest in 2024, the operation included hundreds of thousands of AI-generated songs and billions of automated streams. Spreading the plays across that catalog helped avoid the warning signs that would come from bots all targeting one song.
Smith’s own calculations show the financial logic. In an October 2017 email to himself, he described 52 cloud accounts with 20 bot accounts each. He estimated roughly 661,440 streams a day. At half a cent per stream, his calculation put daily earnings at $3,307.20 and annual earnings above $1.2 million.
An October 2018 email to accomplices made the evasion strategy explicit: Smith wanted a large supply of songs, each attracting relatively few streams, to work around platforms’ anti-fraud policies.
The DOJ illustrated the scale by comparing Smith’s operation with Taylor Swift’s catalog in April 2023. The figures concern YouTube Music family-plan streams—not total listening across every platform. Within that category, Smith’s bot accounts generated far more plays than Swift’s entire catalog.
The DOJ said Smith’s accounts fraudulently streamed his AI-generated music 80.9 million times in this category.
The DOJ put Swift’s catalog at 9.3 million streams in the same month and category.
Before sentencing, Smith’s lawyers sought probation. They argued that officials were making an example of him and that he had not personally pocketed $8 million. Their letter urged the court not to punish one man for the industry’s problems.
Smith expressed remorse and said he had been offered a record-label audio-production job paying $5,000 a month. His lawyers argued that prison would delay his ability to pay the forfeiture.
Prosecutors pressed for incarceration as a deterrent. They told the court Smith had used increasingly deceptive tactics to keep the operation running and had lied to law enforcement when confronted. Their argument was that a financial penalty alone would not adequately punish the sustained fraud or warn others against repeating it.
Loading discussion...
Join the conversation
Explain how you would balance repayment with deterring others.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.