AI songs outstreamed Taylor Swift, but no one was listening
A North Carolina man has been sentenced to 18 months in prison after federal prosecutors said he used AI and thousands of automated accounts to orchestrate one of the biggest music-streaming fraud schemes ever uncovered.
It generated more than $8 million in royalties from songs that were never actually heard by real listeners. The U.S. Attorney’s Office for the Southern District of New York said Michael Smith, 54, pleaded guilty to conspiracy to commit wire fraud and was sentenced on October 6.
Prosecutors described the case as the Justice Department’s “first-ever” criminal prosecution for fraud involving AI-assisted music streaming.
Smith’s lawyers argued he initially believed his conduct was lawful and sought advice from a music lawyer. His legal team had requested probation instead of prison time.
How the Scheme Worked
Streaming services such as Spotify, Apple Music, Amazon Music, and YouTube Music make royalty payments from a shared pool to musicians and songwriters, based on how often songs are streamed.
As federal prosecutors explained, fraudulent streams take that money away from the real-life musicians and songwriters whose songs are actually being played and enjoyed by genuine listeners.
According to the court filings, Smith’s scheme operated from 2017 through 2024 and relied on three components:
- First, he created thousands of fraudulent accounts (called “Bot Accounts“) on streaming platforms including Spotify, Apple Music, Amazon Music, and YouTube Music.
- Next, he used software to get those accounts to continuously stream songs that he had created and, therefore, owned. According to prosecutors, those songs were generally created using AI, which allowed him to build a massive music catalog very quickly. Prosecutors said Smith needed “a TON of songs fast” to avoid the anti-fraud systems that might flag unusually high activity around a small catalog of tracks.
- He then sat back while the royalties—earned by the continuous stream of artificial plays—rolled in. At times, prosecutors said he had as many as 10,000 bot accounts running at the same time. Those accounts streamed his AI-generated songs billions and billions of times, allowing him to collect over $8 million in royalties.
To put the scale of the scheme into perspective, in April 2023, prosecutors said the entire catalog of Taylor Swift received 9.3 million YouTube Music family-plan streams, while Smith’s bot accounts generated 80.9 million streams during the same month.
A Growing Challenge for the Music Industry
AI can generate an enormous amount of music very quickly and very cheaply.
Industry groups say that while this capability can create opportunities for legitimate artists, it can also let bad actors, like Smith, flood streaming platforms with synthetic content designed purely to capture royalties and make money, rather than attract and entertain audiences.
In September, the International Federation of the Phonographic Industry (IFPI) launched the Streaming Integrity Initiative (SII). This has been backed by major labels and distributors including Universal Music Group (UMG), Virgin Music Group, Sony Music Group, and Warner Music Group.
Under the initiative, these labels and distributors have committed to verifying customer identities, screening content for AI-related risks, investigating suspicious activity, and sharing fraud intelligence across the industry.
The organization said that streaming fraud “undermin[es] trust in streaming services and siphon[s] revenues away from the artists, songwriters, record labels, music publishers and others who power the music economy.”
Newsweek has contacted the IFPI for comment and is awaiting a response.
How Streaming Giants Are Fighting AI Music Fraud
Spotify announced in 2025 that, over the past 12 months, it had removed more than 75 million “spammy tracks” and introduced new policies aimed at combating the abusive use of AI in music production and streaming, including stronger action against AI-generated impersonations of artists, new tools to detect and remove spam content, and labels showing when music has been created using AI. The company said harmful AI content can “divert royalties to bad actors.”
Apple Music has also intensified its anti-fraud efforts.
Apple executive Oliver Schusser told The Hollywood Reporter in January that the platform had identified and demonetized roughly 2 billion fraudulent streams during 2025, which equates to around $17 million in royalties that should’ve gone to deserving artists. Apple also increased financial penalties for users attempting to manipulate streaming numbers, with Schusser saying that “increasing the penalties takes the money from people who are cheating and puts it back into the system for those who aren’t.”
Meanwhile, YouTube has strengthened its efforts to address AI-generated and synthetic content by requiring creators to disclose when videos, images, or audio have been generated or significantly altered using AI.
The platform has also rolled out automated systems that can identify AI-generated content and apply labels to videos when creators fail to disclose their use of AI.
Newsweek has contacted Spotify, Apple, and YouTube for comment and is awaiting responses.
What Happens Next
The Smith case shows how easily AI can be used to scale streaming fraud to levels that would have been difficult to achieve manually a few years ago.
The broader challenge for streaming companies will be separating legitimate AI-assisted creativity from industrial-scale content designed solely to exploit the economics of music platforms.
As AI music generation tools continue to improve, the battle is increasingly shifting away from detecting fake songs to identifying fake listeners. The federal case against Smith suggests that regulators, record labels, and streaming platforms view that fight as one of the music industry’s most urgent challenges.
Newsweek’s reporters and editors used Martyn, our AI assistant, to produce this story. Learn more about Martyn here.
Contact Newsweek editors on this story: Ben Kelly and James Debens
