There is a number nobody in the music business wants to say out loud: up to 10 percent of all streams on major platforms may be fraudulent. Some estimates run higher. The bots are not coming. They are already here, they are already working, and in many cases they are already getting paid.

Rolling Stone’s recent investigation into streaming fraud laid out the mechanics clearly. Fake streams are generated through bot farms, networks of devices running automated software that plays tracks on loop without any human ever pressing play. The economics are simple. Streaming platforms pay per play. If you can generate enough plays, you get paid. The fraud is not hidden in dark corners of the internet. It operates at scale, often openly, as a service you can purchase for a few hundred dollars.

The consequences spread in ways that go beyond money. Streaming fraud distorts charts. When a song built on artificial plays climbs the Hot 100 or lands on an algorithmic playlist, it pushes out real music by real artists with real audiences. Those playlist spots matter. They are how new artists get discovered, how catalog tracks find second lives, how emerging artists from small markets break through to wider audiences.

The AI dimension has added a new layer of complexity. AI-generated music, in some cases with no human involvement, has been used alongside bot farms to generate fraudulent royalties at scale. In one widely reported case, a man pleaded guilty to collecting over eight million dollars through exactly this scheme. The mechanics were straightforward: generate tracks, stream them with bots, collect the payout, repeat.

Platforms are aware of the problem. Spotify, Apple Music, and others have fraud detection systems in place. They remove streams, close accounts, and work with distributors to flag suspicious patterns. But the fraud persists because the systems are playing catch-up. The bot operators adapt. When one method gets flagged, they shift to another.

The deeper problem is structural. Streaming royalty systems were designed for a world where a play represented a human choice. They were not designed for a world where plays can be manufactured at low cost. The per-stream royalty model creates a direct incentive for fraud that no amount of detection can fully neutralize as long as the underlying economics remain unchanged.

Some within the industry have proposed moving toward artist-centric models, which allocate royalties based on a listener’s actual consumption patterns rather than distributing the total pool proportionally. This would reduce the reward for generating fake streams against small artists. But it would not eliminate fraud. As long as streams can be bought, some portion of the music economy will be built on them.

For independent artists, the situation is particularly bleak. A fake stream campaign targeting a competitor costs less than a modest ad spend. The barrier to committing fraud is low. The likelihood of being caught is, for smaller operations, relatively modest. The result is a market where legitimate artists compete not just against each other but against manufactured visibility.

The K-pop industry has also been implicated, not necessarily through bots, but through organized fan campaigns that game chart metrics in ways platforms were not designed to account for. This is a different category of manipulation, but it shares the same root: a system built around quantitative measures of popularity that can be gamed, is being gamed.

There is no simple fix. Better detection helps but does not solve the problem. Platform transparency would allow researchers and artists to see more of what is happening, but platforms have little incentive to publicize how much fraud runs through their systems. Regulatory attention is possible but would require a level of technical specificity that most legislators are not equipped to handle.

What is certain is that the current situation is unsustainable as a foundation for a healthy music ecosystem. The money being paid to bot operators is money not being paid to working musicians. The playlist spots being occupied by fraudulent content are spots where real music could be finding audiences. The longer the industry waits to take this seriously, the more normalized the fraud becomes, and the harder it gets to undo.