The fraud was manufactured listening, not AI creation

According to the U.S. Attorney’s Office for the Southern District of New York, Smith created thousands of accounts on music streaming platforms and used software to make those accounts continuously stream music he controlled. He spread the automated activity across thousands of songs to make the pattern harder to detect. AI then gave the operation enough material to scale to hundreds of thousands of songs.

From indictment to guilty plea

The scheme needed both fake listeners and a huge catalog

The important causal chain is therefore large catalog → bot-controlled listening → false appearance of genuine consumer activity → royalty payments. AI helped with the first step. It did not turn legitimate listening into fraud.

Do not turn one fraud case into four bad AI-music rules

• It does not establish that AI music itself is illegal.
• It does not establish one universal AI-labeling rule across every music platform. Disclosure requirements are platform- and use-specific.
• It does not establish that AI-assisted music cannot earn legitimate royalties. The criminal conduct involved manipulated streaming activity.
• It does not make copyrightability the reason Smith’s conduct was fraudulent. Copyright and human-authorship questions are separate from the bot-streaming scheme.

If your concern is AI disclosure on YouTube, use the YouTube AI Rules for Music Creators. If your concern is ownership or copyright, use the AI Music Rights & Ownership Guide.

High output can create the illusion that promotion should scale the same way

AI tools can make it possible to finish more music than one creator could reasonably campaign, contextualize and build an audience around. That is useful only if the creator remains disciplined about what the numbers mean.

Producing 100 tracks does not create 100 audiences. If genuine demand is small, the answer is not to manufacture metrics that make the catalog look bigger than the audience. The better question is which songs deserve focus, what kind of listener responds, and what evidence shows that real people want more.

That is why the repaired JR release system separates making music from building discovery. Start with the AI Music Audience Growth guide and the Music Discovery Strategy.

Legitimate promotion and artificial streaming are not the same thing

Creators can advertise releases, pitch music, collaborate, build communities, send listeners to a release and promote songs through social platforms. The red line is not “promotion.” The problem is activity designed to impersonate genuine listening rather than attract genuine listeners.

For Spotify’s legitimate editorial route, use the Spotify Playlist Pitching 2026 guide.

Six signals that deserve scrutiny

None of these signs alone proves fraud, but they should trigger questions before you hand over money or connect a campaign to your artist profile:

• A guaranteed fixed number of streams or plays.
• A promoter who cannot explain where listeners come from.
• Large listening spikes with little corresponding save, follow, comment or fan activity.
• Traffic concentrated in locations that make no sense for the campaign with no credible explanation.
• Opaque playlist networks presented only as stream-delivery machines.
• Promises focused entirely on the metric rather than reaching an identifiable audience.

Fraud, disclosure and copyright belong in different decision boxes

Streaming fraud: Are the listening signals real or manufactured?

AI disclosure: Does the specific platform or use require disclosure, labeling or another synthetic-content workflow?

Rights and copyright: Do you have the rights you need, and what human-authored material may be protectable?

Keeping those questions separate prevents creators from learning the wrong lesson from a headline. Smith’s case belongs first in the streaming manipulation and creator-trust box.