DOJ Nails AI Streaming Mastermind

Judge holding gavel, hand raised in courtroom. — Photo: Shutterstock

One man turned fake songs and fake listeners into real money—about $8.09 million- until the law caught up.

Story Snapshot

Prosecutors say Michael Smith used artificial intelligence music and bots to siphon royalties.
Smith pleaded guilty to conspiracy to commit wire fraud in federal court.
A judge later sentenced him to 18 months in prison in this first-of-its-kind case.
The scheme shows how fake streams can drain a shared royalty pool that real artists rely on.

The Scheme: AI Songs, Bot Plays, Real Money

Federal prosecutors in New York said Michael Smith built a massive catalog with artificial intelligence tools, uploaded the tracks to streaming platforms, and then drove plays using thousands of automated accounts. The goal was simple: trigger royalty payments by faking demand. The Department of Justice said the operation lasted years and generated about $8.09 million in payouts before platforms raised alarms and investigators moved in. The charge was conspiracy to commit wire fraud, and Smith admitted his role.

Prosecutors described heavy automation across a web of accounts that mimicked normal listening patterns. That camouflage matters because platforms often detect crude fraud. More advanced bots can vary play times, skip rates, locations, and device types. The government’s account pointed to a volume play: flood the system with low-cost content and pump it with traffic that looks human enough to clear filters. That approach depends on scale, not hits. It exploits how streaming platforms count plays and share money.

Why The Money Flow Made It Possible

Streaming services do not pay a fixed price per play. They collect subscription and ad revenue into a pool, then divide that pool by the share of total plays in a period. Every fake stream boosts the fraudster’s slice and shrinks everyone else’s. That structure explains why “no single artist lost a fortune,” yet the system still bled real dollars from real creators. A few million fake streams here or there can feel small, but billions tilt the table.

Researchers and industry reports have warned about this for years. A French National Music Center study defined stream manipulation as fake plays, whether by bots or humans for hire, used to boost income, chart standing, or recommendations. The study also flagged a telltale pattern: streams rising faster than paying users, which makes each stream worth less for honest artists. Smith did not invent the trick; he scaled it with modern tools and crossed a legal line he later admitted.

The Court Outcome And The Defense Pitch

Smith pleaded guilty in March to conspiracy to commit wire fraud. The judge sentenced him to 18 months in prison, plus supervised release, in what prosecutors called the first federal criminal case centered on artificial intelligence-assisted streaming fraud. Prosecutors argued the scheme siphoned $8.09 million, a figure they said already deducted estimated legitimate plays and money Smith cycled back by buying subscriptions used in the scam. That detail undercuts the idea that the loss number was padded.

Smith’s lawyers said he first thought the conduct was lawful after consulting a music attorney, and they asked for no prison time. They argued that using artificial intelligence to make songs is legal, that direct harm to any one artist was hard to show, and that the industry already struggles with widespread stream manipulation. The first point is obvious: artificial intelligence art is not a crime. The issue is the deceit—fake demand, fake listeners, and false statements to keep the money flowing. Wire fraud prosecutions target that intent, and a guilty plea speaks louder than spin.

What Comes Next For Platforms And Artists

Streaming companies say they are getting better at spotting fake plays and punishing bad actors by pulling tracks, holding royalties, and charging penalties. But the core incentive problem remains. As long as one pool pays everyone based on share of plays, fraudsters will try to juice that share. Some scholars propose revenue models that blunt manipulation and reward verified fan engagement, not raw stream counts. That change would align with conservative common sense: protect honest work, punish cheating, and fix systems that invite abuse.

In the first federally prosecuted case of AI-assisted music streaming fraud, a North Carolina man was sentenced to 18 months in prison after using AI-generated songs and thousands of bot accounts to steal more than $8 million in music royalties.

Michael Smith used as many as… pic.twitter.com/rLPlQMQ0bW

— Complex (@Complex) October 12, 2026

Artists and labels can help by watching for odd spikes, click-farm traffic, and sudden playlist adds from shady curators. Distributors should run audits, demand proof of marketing spend, and pause payouts when traffic looks off. Law enforcement has now shown it will treat large-scale streaming deception as a crime, not a prank. When fraud steals from a shared pot, it steals from everyone who plays fair. The Smith case put a price tag on that theft and set a line in bright ink.