Algorithms of Infamy: The Michael Smith AI Streaming Fraud Case and the Collapse of Pro Rata Royalty Models
United States v. Michael Smith
Key Issue
Wire fraud — AI-generated music + bot streaming to divert royalties
Key Takeaways for Practitioners
First major federal prosecution for AI-assisted streaming fraud — Michael Smith diverted over $10 million in royalties using AI-generated 'instant music' and 1,000+ bot accounts.
The scheme generated approximately 661,440 fraudulent streams per day, earning Smith over $1 million annually at the expense of legitimate artists.
The pro rata royalty model's systemic vulnerability: automated streams, no matter how small individually, capture disproportionate market share from the shared royalty pool.
Wire fraud under 18 U.S.C. § 1343 applies cleanly — the scheme defrauded streaming platforms and, derivatively, every legitimate artist whose royalty share was diluted.
The case argues for transition to user-centric royalty models that weight streams by human behavioral patterns — resistant to bot manipulation.
Streaming platforms face increasing pressure to implement fraud detection infrastructure or face secondary liability arguments.
The Music Industry's Reckoning with AI-Automated Fraud
The music industry has faced technological disruption before — cassettes, MP3s, Napster, and the transition to streaming each forced fundamental restructuring. But United States v. Michael Smith represents something qualitatively different: the first federal prosecution establishing that AI-generated content, combined with automated consumption networks, can systematically steal from every legitimate artist in the industry simultaneously.
Smith's scheme was elegant in its simplicity and devastating in its scope. Between 2017 and 2024, he produced hundreds of thousands of algorithmically generated songs — what he called "instant music" — and streamed them through an army of over 1,000 automated bot accounts. The result: approximately 661,440 fraudulent streams per day, generating annual revenues exceeding $1 million diverted from the royalty pool that should have compensated real artists.
The total theft: over $10 million in streaming royalties over seven years.
How the Scheme Worked
The Pro Rata Model's Fatal Flaw
To understand the fraud, you must understand how streaming royalties are distributed. Major platforms — Spotify, Apple Music, Amazon Music — use what is technically known as the pro rata or GLOBALPROP model. Under this system:
- The platform collects all subscription and advertising revenue into a pool
- Each month, the pool is distributed based on market share — the percentage of total global streams each track represents
- Every stream, regardless of who listens, draws from the same pool
The vulnerability is structural: the pool is zero-sum. Every fraudulent stream that captures royalty share is a stream taken from legitimate artists. A sufficiently large volume of automated streams can dilute the value of every real play by a real listener.
Smith did not need to be popular. He needed to be voluminous.
The AI Music Factory
Smith used AI music generation tools to produce content at industrial scale — hundreds of thousands of tracks, algorithmically generated, designed not for human listeners but for automated consumption. The "music" was irrelevant. The streams were the product.
The tracks were uploaded to streaming platforms under Smith's accounts, priced at standard rates, and immediately streamed by his bot network — automated accounts designed to simulate real listeners, distributed across enough IP addresses and device fingerprints to avoid detection thresholds.
The Wire Fraud Theory
The government charged Smith under the federal wire fraud statute, 18 U.S.C. § 1343, which prohibits schemes to defraud using wire communications. The theory has three elements cleanly satisfied:
1. Scheme to defraud: Smith misrepresented to streaming platforms that his streams were genuine human consumption — a material false representation to parties who distribute royalties based on authentic streaming data.
2. Intent to defraud: The scheme's entire design was to divert royalty payments — there is no innocent explanation for 661,440 daily streams from bot accounts.
3. Use of wires: Internet communications and electronic payment transfers constitute "wire" transmissions under § 1343.
The fraud victims are both the streaming platforms (who paid royalties based on falsified data) and, derivatively, every legitimate artist whose royalty share was diluted by Smith's fraudulent market capture.
The Systemic Vulnerability This Case Reveals
Legal Implications for the Music Industry and Streaming Platforms
The pro rata model is structurally vulnerable to AI-assisted fraud. Any distribution system that allocates payment based purely on volume, without weighting for authenticity of consumption, creates an incentive to manufacture volume. The Smith case is not an anomaly — it is a proof of concept that others have already replicated at smaller scales.
User-centric royalty models are the fraud-resistant alternative. Under a user-centric model, each subscriber's payment goes to the artists they actually listen to — not to the global pool. Automated streams from bot accounts, which are not attached to real paying subscribers, would generate zero royalties. Multiple European markets have experimented with user-centric models; the Smith prosecution strengthens the case for adoption in the US.
Streaming platforms face secondary liability pressure. If platforms have fraud detection capabilities they have not deployed, or have detected fraud and failed to act, arguments about secondary liability for royalty theft become stronger. The Smith prosecution puts platforms on notice that the government views streaming fraud as a federal crime — and that notice has legal consequences.
AI music generation is not illegal — fraud is. The Smith case does not criminalize AI-generated music. It criminalizes using any content (AI-generated or human-created) in a scheme to fraudulently divert royalties through fake streams. AI music developers whose tools are used for fraud face no direct criminal liability here — but they should implement terms of service restrictions on fraudulent use and consider fraud-detection provisions in commercial agreements.
What Comes Next
The Smith prosecution is a first precedent, not a complete framework. Several questions remain:
Conspiracy liability: Did streaming platforms, music distributors, or label intermediaries have knowledge of the fraud? The indictment focuses on Smith individually, but conspiracy theories could expand the defendant class.
Civil RICO: Music industry plaintiffs in civil suits may use the Smith prosecution as predicate acts for civil RICO claims against organized streaming fraud rings — a theory that would allow treble damages.
Platform obligations: At what point does a streaming platform's failure to implement fraud detection constitute breach of contract with artists and labels? This question is live in multiple ongoing disputes.
AI detection requirements: Several legislative proposals post-Smith would require streaming platforms to implement AI-content detection and behavioral fraud analysis. The outcome of those proposals will define the industry's regulatory environment for the next decade.
Legal Citation
United States v. Michael Smith, S.D.N.Y. 2024, S.D.N.Y. (September 4, 2024) (Indictment) (Docket No. Filed September 4, 2024)
This analysis is based on publicly available court documents. It does not constitute legal advice.