Spotify said to remove 500K streams of song after suspicious Kalshi bets

Spotify (SPOT) removed more than 500,000 streams from a single song after traders on Kalshi, the prediction-market platform, flagged the activity as suspicious, according to a report. The removal is the first known instance of Kalshi traders acting as an informal fraud-detection network for streaming platforms.

The flagged streams came from a track that showed an unusual spike in plays over a short window. Kalshi users, who bet on streaming counts and chart positions, noticed the anomaly and placed contracts predicting the song would be adjusted downward. Those contracts tipped off Spotify's internal monitoring team, which reviewed the data and confirmed the streams were generated by automated scripts rather than real listeners.

The incident ties prediction markets to entertainment fraud in a new way. Kalshi launched music-streaming prediction contracts earlier this year, letting traders wager on metrics such as weekly Spotify stream counts and Billboard chart positions. Because those markets attract participants with deep knowledge of the streaming ecosystem, artificial activity often surfaces there before platforms catch it. Kalshi is regulated by the Commodity Futures Trading Commission and offers contracts on a range of real-world events. Its music-streaming contracts are among its newer offerings.

For Spotify, the episode is a reminder that its 500-million-user base remains vulnerable to bot-driven inflation. The company has spent years building anti-fraud technology, including machine-learning models that flag velocity anomalies and account-creation patterns. The Kalshi case is an example of external signals filling gaps in internal detection, especially when the fraud targets a specific song rather than a broad playlist.

The removal of 500,000 streams is a small fraction of Spotify's total daily streams, which number in the billions. The method of detection, via a prediction market, is novel. The financial impact of the removed streams is small. At Spotify's average per-stream payout of about $0.004, the royalties amount to roughly $2,000. The reputational risk is larger. Labels and artists already dispute payout calculations, and a high-profile manipulation case could intensify calls for more transparent streaming audits. Major record groups have pushed for better fraud reporting in recent licensing negotiations. The music industry has long complained about streaming fraud. The Recording Industry Association of America has called for better detection methods.

Kalshi declined to comment on the specific contracts. A spokesperson for Spotify said the company removes fraudulent streams "on an ongoing basis" and declined to elaborate on the song in question.

The broader industry effect may be felt in how platforms treat external data. If Kalshi's prediction markets continue to surface manipulation before official detections, streaming companies could face pressure to integrate such signals into their own systems. The Songwriters Guild of America has already asked the Copyright Office to study whether streaming platforms are doing enough to prevent fraud. For Kalshi, the incident is a validation of its prediction-market model. The platform has been seeking to expand beyond financial and political events into entertainment and culture. The Spotify case could attract more traders to its music contracts.

The incident is the clearest example yet of prediction markets functioning as an early-warning system for a non-financial metric. Whether that role expands depends on how much trust labels and platforms place in a betting market's signals. For now, the episode shows that traders with domain knowledge can spot anomalies faster than internal systems.