Streaming fraud is the use of fake listeners, often automated bots, to inflate play counts and take royalty money from shared pools. Ars Technica reports that one such scheme, which relied on AI-generated songs, ended in an 18-month prison sentence.
Defining streaming fraud
Streaming fraud, also called artificial streaming, is any attempt to generate plays on a music service that do not come from real people choosing to listen. The aim is usually financial. Streaming platforms pay royalties to rights holders according to how often their recordings are played. If someone can produce plays without real listeners, they can claim money that would otherwise go to genuine artists, songwriters and labels.
The most common methods are automated programs, known as bots, which log in to accounts and play tracks in a loop. Other methods include paid “click farms” running many physical devices, compromised user accounts, and playlist services that promise exposure but deliver fake plays. A newer variation uses generative artificial intelligence to produce large catalogues of songs at almost no cost. Bots then stream those songs.
Ars Technica reports that a man has been sentenced to 18 months in prison for a scheme of this kind. According to the publication, it used about 10,000 bots and AI-generated songs and took roughly $8 million in streaming royalties. This article covers the general practice rather than the individual case.
Origins of the problem
Streaming fraud grew alongside the subscription and advertising-funded music services that replaced downloads and physical sales as the main way people pay for recorded music. Under these models, payment is tied to play counts rather than to a one-off purchase. That link between plays and money is what makes fake plays worth producing.
Early schemes mostly involved inflating the numbers of real tracks, sometimes by artists or promoters trying to look more popular. The incentive changed as digital distribution became open to anyone. Independent distributors let almost anyone upload music to major platforms, which lowered the barrier to entry for honest musicians and fraudsters alike. Cheap cloud computing, plentiful stolen or fabricated account credentials, and later AI music generators each made large-scale fraud easier and cheaper to run.
The exact scale of streaming fraud across the industry is not reliably known. Estimates differ, and platforms rarely publish detailed figures about what they detect.
How schemes operate today
A typical modern scheme has three parts: content, accounts and distribution of plays.
Content. Fraudsters need tracks they control so that the royalties flow to them. Generative AI tools can produce thousands of short, plausible-sounding songs quickly. These are released under many invented artist names, so that no single artist or track draws attention.
Accounts. Bots need accounts to stream from. These may be paid subscriptions bought in bulk, free accounts, or accounts taken over from real users. Paid accounts often generate more royalty value per play, which can make them attractive despite the cost.
Spreading the plays. The central technique is dilution. A single song with millions of plays from a small set of accounts is easy to spot. Spreading modest play counts across a very large catalogue makes each track look like an ordinary low-level release. In aggregate, though, the plays can add up to substantial sums. This is how a scheme built on obscure, machine-made songs can rival the total streams of the most popular artists.
Most major services pay out through a pro-rata model. Subscription and advertising revenue goes into a pool, which is then divided according to each rights holder’s share of total plays. Fraudulent plays therefore do not create new money. They take a larger slice of a fixed pool, which reduces what legitimate rights holders receive.
Platforms and distributors use detection systems to counter this. These look for unusual listening patterns, clusters of accounts behaving identically, suspicious network origins and abnormal ratios between plays and other engagement. Some services have also changed their royalty rules to make low-value fraud less profitable, and some penalise distributors whose uploads show signs of manipulation. The internal details of these systems are generally not published.
Common misconceptions
“It is a victimless trick against big tech companies.” Under pro-rata payment, the money comes out of the shared royalty pool. Real artists and songwriters are the ones who lose income.
“AI music is the fraud.” Producing music with AI tools is not in itself illegal or fraudulent. The fraud lies in generating fake listening to collect royalties. AI-generated tracks are simply a cheap source of content for such schemes.
“Buying streams is just marketing.” Many services that sell plays or playlist placements deliver bot traffic. Their terms of service usually prohibit this, and artists who use these services can have tracks removed or royalties withheld, even if they did not fully understand what they were buying.
“Only huge play counts get noticed.” Detection looks at behaviour, not just volume. Schemes spread thinly across many tracks still leave patterns in accounts, devices and timing.
“It is only a terms-of-service problem.” The case reported by Ars Technica shows that streaming fraud can lead to criminal prosecution and imprisonment, not only account bans.
Where to look next
People who want to understand the subject further can start with the published artist and distributor policies of the major streaming services. These usually explain what counts as artificial streaming and what penalties apply. Music industry trade bodies and collecting societies publish material on royalty distribution and fraud prevention. Court records and official prosecution announcements give the most reliable account of individual cases, including the charges and outcomes. Specialist technology and music business journalism, such as the Ars Technica report behind this article, tracks how schemes and countermeasures develop. Anyone researching AI-generated music more broadly should keep the creative and legal questions about AI composition separate from the narrower question of fraudulent streaming.
Frequently asked questions
What is streaming fraud?
Streaming fraud is the use of fake listening activity to inflate play counts on music platforms, usually so that someone can collect royalties. It often relies on bots that play tracks automatically from many accounts. Because most services split a shared revenue pool according to share of total plays, fraudulent streams divert money away from genuine artists, songwriters and labels rather than creating new revenue.
How do bots make money from music streaming?
Bots log in to streaming accounts and play tracks that the fraudster controls, so the royalties for those plays go to the fraudster. To avoid detection, schemes often spread plays across a large number of songs and invented artist names. Each track then looks unremarkable, but the combined plays can generate significant payments from the platform’s royalty pool.
Is it illegal to upload AI-generated music to streaming services?
Uploading AI-generated music is not in itself fraud. Rules depend on each platform’s policies and on copyright questions that are still developing in many jurisdictions. What becomes fraudulent is generating fake plays to collect royalties. AI tools make such schemes cheaper because they can produce large catalogues of tracks quickly, but the deception is in the artificial listening.
Who loses money when streams are faked?
Under the pro-rata model used by most major services, subscription and advertising revenue goes into a common pool. That pool is divided by each rights holder’s share of total plays. Fake plays increase the fraudster’s share and reduce everyone else’s, so legitimate musicians, songwriters, publishers and labels all receive less than they otherwise would.
Can streaming fraud lead to prison?
Yes. Ars Technica reports that a man received an 18-month prison sentence for a scheme that used about 10,000 bots and AI-generated songs to take roughly $8 million in royalties. Beyond criminal cases, platforms can remove tracks, withhold royalties and penalise distributors linked to artificial streaming, depending on their published terms.
Sources and further reading
- Ars Technica: technology policy reporting on the sentencing in an AI-assisted streaming royalty fraud scheme
- Streaming platform artist and distributor policies: published rules on artificial streaming and associated penalties
- Music industry trade bodies and collecting societies: material on royalty distribution and anti-fraud efforts
- Court and prosecutorial records: official documents on charges and outcomes in streaming fraud cases
Surfaced from the rss:arstechnica signal “music streaming royalty fraud”. AI-assisted draft, editorially reviewed.
