A 54-year-old North Carolina man was sentenced to 18 months in federal prison for orchestrating a seven-year scheme that used thousands of automated bot accounts and artificially generated songs to siphon millions of dollars in royalties from music streaming platforms — a case the Justice Department says is the first of its kind in the United States.

Michael Smith, who pleaded guilty before sentencing, was charged in 2024 and sentenced this week, according to a Department of Justice press release issued Tuesday. Prosecutors described a long-running operation in which Smith created AI-generated tracks, distributed them across streaming services, and then used a network of roughly 10,000 bot accounts to play those tracks on repeat — inflating his stream counts and, by extension, his payouts.

The mechanics of a modern royalty heist

The scheme exploits the basic arithmetic of streaming economics. Platforms pay rights holders a fractional amount per stream — typically a fraction of a cent — meaning that meaningful revenue requires enormous volume. Independent artists often struggle to reach that scale legitimately. Bot-driven "stream farming," or fraudulent streaming, shortcuts the process by manufacturing the volume directly.

Smith's operation added a newer twist. Rather than simply looping a handful of tracks, he used artificial intelligence to generate songs at scale, filling catalogs quickly and cheaply. That combination — synthetic music plus automated playback — is what distinguishes this prosecution from earlier streaming-fraud cases involving human-created tracks and manually managed bot farms.

Reporting on the case diverged slightly on the scale of the losses. Some outlets pegged the fraud at approximately $8 million, while others placed it at $10 million. The discrepancy reflects the difficulty of calculating damages in a system where royalties accrue continuously across multiple platforms, distributors, and payment cycles over seven years. Ars Technica noted that Smith's operation ran undetected for those seven years before investigators caught up with him.

The Justice Department described the prosecution as the first time an American has been criminally charged with AI-assisted streaming fraud — a signal that federal prosecutors intend to treat synthetic-media manipulation of royalty systems as a serious financial crime.

Outstreaming Taylor Swift

Several accounts of the case framed Smith's success in terms of what his bot network managed to out-perform: the artificially inflated tracks were pushed past the streaming totals of major artists, including Taylor Swift. The framing is vivid, but it also captures something structural. Streaming charts, playlist algorithms, and payout thresholds all respond to raw play counts. When those counts can be manufactured, the same signals that reward legitimate viral hits can be hijacked.

That vulnerability is not theoretical. Platforms have spent years tightening detection around anomalous listening patterns — sudden spikes, low-skip replay loops, geographic clustering of plays, and accounts that only ever stream a narrow catalog. The Smith case suggests those defenses can be defeated, at least for a time, when the fraudulent catalog itself is generated to look like ordinary content rather than a loop of the same file.

Why the AI angle matters

Generative audio tools have collapsed the cost of producing passable music. Where a fraudster once needed a roster of real artists to fill a catalog, an AI pipeline can produce thousands of tracks in hours. Distributors, which sit between artists and platforms, generally require only minimal vetting to upload content, and the resulting catalog volume makes manual review impractical.

That asymmetry — cheap generation on one side, expensive verification on the other — is the core problem the industry now faces. Rights holders, including major labels and independent artists alike, have long argued that fraudulent streaming dilutes the royalty pool, meaning legitimate musicians are literally paid less because bots are paid more. The Music Fraud Alliance estimates the practice drains hundreds of millions of dollars annually from the industry, though precise figures remain contested.

Platforms have responded with policy changes, including minimum play thresholds before a track earns royalties and stricter distributor requirements. None of those measures fully addresses AI-generated catalogs backed by distributed bot networks.

What the sentence signals

An 18-month sentence is relatively modest compared with the alleged scale of the fraud, and legal observers note that the case's most significant impact may be precedential rather than punitive. By charging Smith under wire fraud and related statutes rather than treating the conduct as a civil matter or an obscure terms-of-service violation, federal prosecutors have established that AI-assisted streaming manipulation can be prosecuted as a financial crime.

The case also raises unresolved questions. How should platforms distinguish between genuinely popular AI-generated music and fraudulently inflated AI-generated music? What liability, if any, falls on the distributors that uploaded the tracks? And as generative audio improves, will detection systems be able to keep pace with the volume of content they must screen?

For now, the Smith case stands as a marker: the first American to go to prison for using artificial intelligence to game the music streaming economy — and almost certainly not the last.