Copyright Office opens streaming fraud inquiry with AI music in its sights

The US Copyright Office is taking comments until November 23 on how streaming fraud works and which distributor policies might stop it, including customer verification, royalty clawback and content removal. If you release your own edits or originals through a distributor, those are the policies under review. Reply comments are due December 21, and the notice of inquiry went into the Federal Register on October 7.

The booking side is in scope too. The notice names effects on playlist placement and on the streaming data that artists and venues rely on when planning tours as secondary harms, and it asks how fraud affects live music, including venue selection and ticket sales.

The push came from Rep. Scott Fitzgerald, a Wisconsin Republican on the House Judiciary Committee. His May 21 letter asked the Office to look at AI’s role in streaming fraud. He was careful to say there is “nothing inherently wrong” with AI as a creative tool. His complaint is about the pot of money: “When AI-generated outputs compete against human-made works, they divert money from a finite royalty pool that would otherwise be distributed to actual rights holders,” he wrote. The letter leans on Deezer’s numbers: up to 85% of streams on fully AI-generated tracks in 2025 were fraudulent, and by June more than half of daily uploads were AI-generated, close to 90,000 tracks a day.

What the notice asks about distributors

Michael Smith, of Cornelius, North Carolina, was sentenced to 18 months in prison the day before the notice was published. He ran bot accounts that streamed hundreds of thousands of AI-generated songs. The notice repeats the $10 million in royalties alleged at his 2024 indictment, but the judge ordered a forfeiture of $8,091,843.64, the loss figure both sides agreed on.

Smith is the easy case. The notice gets more interesting where it asks about ordinary plumbing. Among its 10 subjects: “What industry policies or practices may be inadvertently contributing to streaming fraud?” It wants views on trusted distributors, content verification, customer verification, royalty clawback, fines and content removal. It asks about a shared database of known fraud, standardized metadata, and labeling content that is ineligible for statutory royalties. It also floats “no-fault injunctive relief” and names two bills, the American Copyright Protection Act of 2026 and the DEFEND IP Act.

None of that is policy yet. If these ideas are adopted, though, expect distributors to ask for more verification from the people who upload, and to claw back royalties paid on streams they flag as fraudulent.

The industry is moving on its own track. IFPI launched the voluntary Streaming Integrity Initiative on September 14, with five baseline anti-fraud practices for distributors, including Know Your Customer checks and intelligence sharing. Twenty-four companies and trade bodies signed at launch, the three majors among them. The notice references the initiative, and cites DistroKid’s and TuneCore’s guidance on artificial streaming as examples of distributors that penalize fraud.

Organized crime and what comes next

There is also an organized-crime thread. The notice calls a 2023 Swedish report “alarming”: it found criminal networks using fake streams to launder money. NMPA chief David Israelite called streaming fraud “a national security issue” at the trade body’s June 10 annual meeting. State and local prosecution is on the agenda too.

Don’t expect instant law. An earlier inquiry prompted by a Fitzgerald letter, into performance rights organizations, drew more than 5,000 comments. The Office’s November 2025 letter to Congress summarizing them “does not contain new policy recommendations,” in the Office’s own words.

If you release music, read your distributor’s artificial-streaming policy now, and keep records of where your plays come from.