September 5th, 2024

The Bands and the Fans Were Fake. The $10M Was Real

North Carolina musician Michael Smith faces fraud charges for using AI to create fake music, allegedly earning $10 million by manipulating streaming services over seven years, risking 20 years in prison per charge.

Read original articleLink Icon
The Bands and the Fans Were Fake. The $10M Was Real

A North Carolina musician, Michael Smith, has been charged with fraud for allegedly using artificial intelligence to create fake music and manipulate streaming services to earn $10 million in royalties. Federal prosecutors claim that Smith generated hundreds of thousands of fictitious songs attributed to non-existent bands, which he then streamed using bots to simulate real listeners. This scheme reportedly spanned seven years, during which Smith created thousands of fake streaming accounts and used software to loop his music, giving the illusion of genuine engagement. The indictment details how he initially uploaded his own music but later expanded his catalog by collaborating with an AI music company, producing a vast array of bogus songs. By 2019, Smith was reportedly earning around $110,000 monthly from this fraudulent activity. He faces charges including wire fraud and money laundering, with potential sentences of up to 20 years for each count. This case marks the first criminal prosecution related to streaming manipulation by the U.S. Attorney's Office in Manhattan, highlighting the growing concern over digital music fraud as the industry increasingly relies on streaming metrics for success.

- Michael Smith is charged with using AI to create fake music for fraudulent royalties.

- He allegedly earned $10 million over seven years through streaming manipulation.

- Smith created thousands of fake accounts and used bots to simulate listeners.

- The case is the first of its kind prosecuted by the U.S. Attorney's Office in Manhattan.

- If convicted, Smith faces up to 20 years in prison for each charge.

Link Icon 6 comments
By @mistersquid - 8 months
> The man, Michael Smith, 52, was accused in a federal indictment unsealed on Wednesday of stealing royalty payments from digital streaming platforms for seven years. Mr. Smith, a flesh-and-blood musician, produced A.I.-generated music and played it billions of times using bots he had programmed, according to the indictment.

Fascinating. Comments in the OP (presently 18) question whether Smith’s scheme is illegal or merely an unethical exploitation of a loophole.

One commenter raises the point that humans are coerced into producing meaningless content only to be exploited by large corporations. Other commenters question whether streams need be listened to by humans at all.

What if the stream is being “listened” to by an algorithm or as training data for a neural network? What if the stream is being played to people engaged in other tasks or to pre-verbal children, to people in their sleep?

What qualifies a stream to be non-fraudulent?

By @cthalupa - 8 months
Reminds me of https://en.wikipedia.org/wiki/Sleepify

> Sleepify is an album by the American funk band Vulfpeck, released March 2014. The release consists solely of ten roughly 30-second-long tracks of silence. The album was made available on the music streaming service Spotify, where the band encouraged consumers to play the album on a loop while they slept. In turn, royalties from the playing of each track on the "album" were to be used to crowdfund a free concert tour by the band.

They basically took advantage of how Spotify royalty calculation worked and ended up making $20k off of it, which did then get used to fund a small free admission tour of the US.

By @TomMasz - 8 months
Fake bands are not a new thing. Of course, that usually meant a real vocalist and real (usually studio) musicians but those songs were still sold for money. If this guy produced songs that people listened is it any different than what came before?

The fake listeners, though, aren't the same thing. Bots don't buy products.

By @spudlyo - 8 months
By @samspenc - 8 months