A North Carolina man has pleaded guilty to defrauding music streaming platforms and their fellow musicians of millions of dollars in royalties by flooding the service with thousands of AI-generated songs and using automated “bots” to artificially increase the number of listens to billions.
As part of an agreement with federal prosecutors in the Southern District of New York, 52-year-old Michael Smith pleaded guilty Friday to conspiracy to commit wire fraud.
The Cornelius, North Carolina, resident’s lawsuit is one of the first to successfully prosecute AI-related fraud cases in the music industry, which is at risk of disrupting streaming services and depriving legitimate human musicians and copyright holders of revenue from fake music.
“Michael Smith used artificial intelligence to generate thousands of fake songs and streamed those fake songs billions of times,” U.S. attorney Jay Clayton said in a statement.
“While the songs and listeners were fake, the millions of dollars Smith stole were real. Millions of dollars in royalties that Smith misappropriated from real, legitimate artists and rights holders. Smith’s brazen scheme ended after he was convicted of a federal crime for fraud using AI.”
Smith was indicted in September 2024 for amassing 661,440 daily streams from 2017 to 2024 and fraudulently obtaining $1,027,128 in annual copyright royalties.
At the time, U.S. attorney Damian Williams said the defendants had stolen “millions of dollars in royalties owed to musicians, songwriters, and other rights owners whose songs were legally streamed,” adding, “It’s time for Mr. Smith to face his music.”
As one X commentator using the handle Tuki pointed out after the plea deal was announced, Smith was making $1.2 million a year “on music that humans have never actually heard,” taking advantage of “AI creating music and audiences.” Musicians and the music industry currently have to “fight with songs that don’t exist being heard by people that don’t exist,” user X added.
Under the terms of his plea agreement, Mr. Smith now faces up to five years in prison and forfeiture of $8,091,843.64 when sentenced in July.
The case against Smith highlights the growing problem that the music industry, which has largely recovered from the Napster music piracy era of the early 2000s, faces AI-based threats to revenue from music streaming platforms such as Amazon Music, Apple Music, Spotify and YouTube Music.
With the exception of a few big stars, musicians have long complained that they only earn enough to make a living, and their business model pays them out of a pool of money proportional to their income. But AI-generated music, and AI-related plans to drive playback, divert funds from musicians and songwriters whose songs are legally streamed by real consumers.
The UK government recently abandoned plans to allow AI companies to use copyrighted works without permission, a proposal strongly opposed by thousands of artists including Elton John, Dua Lipa and Paul McCartney.
The issue of generated AI music has drawn attention to Suno, a company with 2 million subscribers that can provide users with AI-generated music that disrupts the act of creation.
French streaming service Deezer suggests that 97% of people can’t tell the difference between music created by humans and music created by AI. This includes 60,000 fully AI-generated tracks delivered to the service every day.
According to US trade publication Billboard, Suno generates 7 million songs a day, which is equivalent to the streamer’s entire fortnightly music catalog. Much of the output bears a fair resemblance to existing music composed by humans, but like most AI creations, it comes across as mass-produced, devoid of artistic risk or depth.
Suno CEO Paul Sinclair told Billboard in early March that he was conflicted. “Really, I’m conflicted every day,” he is quoted as saying. “This is complicated…I want to make sure that the beauty of art and music and the ability to build a career around it remains for an entire future generation.”
