Music companies and distributors are stepping up efforts to stop AI-generated music from manipulating streaming platforms and diverting royalty payments from genuine artists and songwriters.
The International Federation of the Phonographic Industry (IFPI), which represents the global recorded music industry, announced new guidelines under which distributors will take stronger measures to prevent fraudulent music from reaching streaming platforms, the Financial Times reported.
The move comes as generative artificial intelligence makes it easier to create large volumes of music and use automated accounts to artificially increase stream counts.
According to the news report, industry executives estimate that nearly 10 per cent of music streams are now fraudulent, with AI-generated music accounting for a significant share of such activity.
How AI is being used to manipulate streams
Streaming fraud typically involves creating large numbers of tracks and then using networks of automated accounts, or bots, to play them repeatedly. Generative AI has made the first part of this process much easier. Instead of spending time and money producing individual songs, fraudsters can generate thousands of tracks using simple prompts. They can then distribute the songs across streaming platforms and use bots to inflate their play counts.
Because royalties are distributed from a pool based partly on the number of streams, fake plays can reduce the share available to artists whose music is being played by genuine listeners.
French streaming platform Deezer said more than half of the new music uploaded to its service in July was AI-generated, at about 90,000 tracks a month. It also found that as many as 85 per cent of streams of fully AI-generated tracks in 2025 were fraudulent, the news report said.
Michael Smith case shows scale of fraud
A US criminal case has highlighted how AI can be used to scale up streaming fraud. Michael Smith, a 54-year-old North Carolina man, pleaded guilty in March 2026 to conspiracy to commit wire fraud in what the US Department of Justice described as an AI-assisted music streaming fraud scheme.
According to the department, Smith created hundreds of thousands of AI-generated songs and thousands of accounts on streaming services. He then used software to make automated accounts repeatedly stream the songs, generating billions of fake plays.
The scheme allowed Smith to obtain more than $8 million in royalties, the Justice Department said. He agreed to forfeit $8.09 million and faces a maximum possible prison sentence of five years.
The case shows how AI can make streaming fraud much larger than traditional schemes by allowing fraudsters to create vast catalogues of songs at very low cost.
Spotify moves to label AI artist profiles
Streaming platforms are also changing how they identify and handle AI-generated music. Spotify plans to introduce an “AI Persona” label for profiles where the artist’s identity appears to be AI-generated, according to TechCrunch. The labels are due to appear from September and will be visible on artist profiles, search results and track listings.
Spotify said it will not depend only on artists identifying themselves. It will also review profiles whose names and images appear to represent photorealistic AI-generated identities.
Music from these AI Persona profiles will generally be excluded from Spotify’s editorial and algorithmic recommendations, as well as personalised recommendations, unless a user already follows the profile, the news report said.
Spotify has also introduced AI-related music credits and has said unauthorised AI voice clones and deepfakes are not allowed on its platform.
YouTube faces its own AI content challenge
YouTube has also been working to improve transparency around AI-generated content. In May, the platform said it was introducing more prominent AI disclosure labels and new systems to automatically detect some AI-generated content. Labels for long-form videos appear below the video player, while Shorts can carry the label directly on the video.
YouTube said creators still have to disclose realistic AI use, but its internal systems can automatically apply labels when significant AI use is detected. Content created using YouTube’s own AI tools or carrying verified C2PA information showing it is fully AI-generated can receive labels that cannot be removed.
The challenge for YouTube, as with music-streaming platforms, is to distinguish legitimate use of AI as a creative tool from mass-produced content designed to exploit recommendation and monetisation systems.
Industry seeks common rules
The new Streaming Integrity Initiative supported by IFPI brings together more than a dozen labels and distributors, including Sony Music, Universal Music, Warner Music and The Orchard.
Under the guidelines, distributors will be expected to verify rights ownership and customer identities, check content for copyright infringement and fraudulent activity, assess AI-related risks and take action against repeat offenders. They will also share information about fraudulent activity, the Financial Times reported.