Global recorded music revenue reached US$29.6 billion in 2024. The streaming economy, a cornerstone of the wider music industry, exceeded US$20 billion in revenue for the first time the same year. It is also a key target for criminals looking to perpetrate streaming fraud.
Using armies of bots or forming entire streaming farms, fraudsters artificially inflate streaming numbers, diverting billions of dollars from the finite royalty pool – funds that should be allocated to right holders such as music creators, artists, labels, or publishers – into their own bank accounts. Music streaming platforms distribute royalties based on play counts but by manipulating the system, criminals can undermine business models across the industry.
This is not new. Streaming fraud has existed for as long as streaming platforms. However, the explosion of AI technology has poured rocket fuel on this long-smoldering issue, revolutionizing the way fraudsters operate and supercharging their ability to evade detection.
Illegitimate royalties generated by artificial streaming
In the past, bad actors would upload a relatively small number of tracks to streaming platforms and have automated bots play the content repeatedly to generate fraudulent royalty payments. The problem for the would-be thieves, though, is that this approach is easy to spot; unknown songs by unknown artists suddenly garnering millions of streams raises obvious red flags. AI has flipped the paradigm on its head.
Now, fraudsters use AI song generators to flood streaming platforms with millions of fake songs and stream each one just a few thousand times – enough to generate royalties from each track but not enough to arouse suspicion and detection.
Deezer estimates that 18 percent of the content uploaded to the platform every day is AI-generated
“AI is the ultimate enabler” of streaming fraud, Melissa Morgia, Director of Global Content Protection and Enforcement at the International Federation of the Phonographic Industry (IFPI), told a panel on the sidelines of the seventeenth session of WIPO’s Advisory Committee on Enforcement (ACE) in February 2025, because it allows bad actors to “stay under the radar but still operate at a sufficient scale that their activities are lucrative”.
Michael Smith AI music streaming fraud lawsuit
The recent case of North Carolina musician Michael Smith is emblematic of this new form of artificial streaming fraud. Smith allegedly extracted more than US$10 million in royalty payments from a host of streaming platforms by uploading hundreds of thousands of AI-generated songs and using bots to play each one a smaller number of times.
Bad actors are using AI not only to generate audio content but also to create and manage the bots used to stream the content. There are even businesses that boldly advertise streaming fraud as a service, highlighting their use of AI to spoof digital identities en masse and “bypass anti-fraud systems” employed by the likes of Spotify, Apple Music and Deezer. Companies pushing the use of bots frame streaming fraud as a valid way for musicians to grow their brand but conspicuously avoid any mention of the damage it causes across the music industry.
Financial impacts of artificial streams on the music industry
The most obvious and direct harm is financial. Streaming platforms have a finite revenue pool from which they can pay royalties and every time a bad actor successfully extracts fraudulent payments, there is less revenue to share with artists, labels and publishers.
Everyone in the value chain loses out on a material amount of revenue on an annual basis
In April 2025, streaming platform Deezer estimated that 18 percent of the content uploaded to the platform every day is AI generated. That’s about 20,000 tracks.
Morgan Hayduk, Co-CEO and Co-Founder of Beatdapp, a service that identifies streaming fraud and tracks missing royalties, believes this figure largely holds true across the music streaming ecosystem, which would mean massive financial losses for the industry.
“Every point of market share is worth a couple hundred million US dollars today,” Hayduk tells WIPO Magazine. “So we’re talking about a billion dollars minimum – that’s a billion dollars being taken out of a finite pool of royalties and everyone in the value chain losing out on a material amount of revenue on an annual basis.”
Knock-on effects of fake streams on real artists
Beyond loss of revenue, streaming fraud has a number of knock-on effects. Every time a song’s play count is manipulated, it skews its platform’s recommendation algorithm and makes it more difficult for real artists to get their music heard. It also distorts the consumer data that artists increasingly rely on to plan tours and promotional campaigns and decreases the window of opportunity artists have to get a toehold in the music business.
As David Sandler, Warner Music Group’s Vice President of Global Content Protection, put it at the panel: “[Streaming fraud] is impacting artists you’ve never heard of because we don’t have a chance to bring them to market. Our company invests a tremendous amount of money, time and energy in discovering new artists, signing new artists and developing their careers. Every dollar we spend to fight fraud is a dollar we can’t spend discovering new artists.”
Streaming trackers for fraud prevention and detection
As the threat of streaming fraud increases, so too do the industry’s mitigation efforts. Using many of the same technologies employed by fraudsters, stakeholders are developing new detection tools to identify AI-generated content and manipulated streams.
The legal tools to take action globally are there
“AI can do good things too,” says Thibault Roucou, Deezer’s Royalties and Reporting Director. “We have been using AI to fight this since 2017, to detect user behavior that is fraudulent and content that is suspicious.”
In addition to its AI-based solutions, Deezer has introduced a new artist-centric remuneration model with an innovative approach to combating streaming fraud. When calculating royalty payments, Deezer caps its users at 1,000 streams – if a single user surpasses this limit, they are still able to listen to music but generate royalties at a much lower rate.
“It means that, with only one account, you cannot generate thousands and thousands of streams and redirect royalties,” says Roucou. “It’s very useful for combating bots who are just looping content endlessly.”
Joining forces to crack down on streaming farms
Despite these promising developments, the solution to this issue lies beyond the actions of any single company or even any single government. The networks that enable these fraudulent practices operate worldwide, which means any mitigation efforts must be similarly broad.
On the policing and enforcement side, IFPI’s Melissa Morgia notes that many of the necessary mechanisms are already in place; the challenge lies in helping local authorities familiarize themselves with the legal issues and in facilitating communication between music industry stakeholders and the jurisdictions where fraud networks are operating.
“The legal tools to take action globally are there,” says Morgia. “It’s just a matter of implementation.”
For industry stakeholders, it is paramount to share data on the rates, types and methods of detected fraudulent activity and take collective action on this issue. In 2023, global music companies including Spotify, SoundCloud and TuneCore united to form the Music Fights Fraud Alliance. It works with the National Cyber-Forensics and Training Alliance and represents the most coordinated industry action to date and is a foundational step towards combating the issue.
As we move into the age of AI, threats to intellectual property (IP) are evolving and multiplying at an astonishing rate. The sophistication and scale of attacks on IP holders and their royalty payments will only increase over the coming years. It is essential that stakeholders from across the industry work together with public institutions and global organizations to fight fraud.
As Sandler points out: “This is a global problem – it crosses borders, it crosses streaming platforms, and we need a unified effort.”
About the author
Clovis McEvoy has a Master’s degree in contemporary music composition and lectured at the University of Auckland, NZ, teaching music production, sound design and composing for film and game. He’s a regular contributor to a number of online publications, including MusicTech, MusicRadar and Future Music. He’s also an award-winning composer, sound artist and founding member of the multidisciplinary group Rent Collective.
Find all latest articles with a focus on music in this special edition of WIPO Magazine.
