Kalshi Settled $3.3M Spotify Market After Fraud Warning

A Kalshi trader warned the exchange that streaming data behind a $3.32 million market on which artists would reach No. 1 on Spotify’s U.S. daily chart appeared manipulated. Kalshi settled the market minutes after its head of enforcement replied that only Spotify could verify the streams.

Trader Caleb Davies says he first contacted Kalshi CEO Tarek Mansour after detecting abnormal U.S.-only streaming increases in several songs as June neared its end. Mansour referred Davies to Robert DeNault, Kalshi’s head of enforcement and legal counsel. Davies provided his analysis and asked Kalshi to hold payout until an investigation was complete. DeNault wrote the exchange was examining the information, noted the main point of contention involved another platform, and added that ‘only Spotify can verify’ whether the listening data reflected genuine streams or manipulation.

The contested outcome was Malcolm Todd’s song ‘Earrings,’ which rose roughly 70% in a single day and was listed as No. 1 on Spotify’s June 29 U.S. daily chart. Davies calculated the Sunday-to-Monday spike as an 11.24 sigma event and said that of about 200 songs charting on both days, 195 declined while four rose, including two he had flagged. Open interest in the market bracket backing ‘Earrings’ rose from about $2,000 to more than $70,000 in the days before the surge. Traders had priced ‘Earrings’ at roughly 2.5% in the prior week, implying about a 20-fold return for holders.

Kalshi’s trading ledger shows $3,320,917 in volume for the June market. DeNault’s reply to Davies came six hours after the warning; Kalshi settled the market minutes after sending that reply. Davies published the correspondence and recorded a loss of roughly $4,500 on the settlement. He says Kalshi treated this payout with greater urgency than other contested results, noting an Olivia Rodrigo outcome went unpaid for three days after he flagged it and a Michael Jackson payout took over a week.

The day after Kalshi settled, Spotify removed 523,000 streams from the cumulative total for ‘Earrings.’ Removing those streams would have placed the song fourth rather than first for June 29. Spotify does not revise its published daily charts retroactively, so the chart Kalshi paid against remains unchanged. Spotify has not identified who generated the removed streams or shown the activity was intended to influence a prediction market. There is no public evidence linking Todd or his team to the removed streams.

Following the dispute, Spotify asked platforms to remove its branding and to state they do not have partnerships with the streaming service. Kalshi removed the Spotify logo and revised wording that had suggested Spotify verified its markets. A company spokesperson, Elisabeth Diana, wrote that Kalshi was ‘in touch with Spotify’ and actively investigating. Kalshi has not published the results of an internal review, reversed the settlement, or announced reimbursements. Davies says no one at Kalshi has contacted him since Spotify confirmed the fraudulent streams.

Kalshi has stopped listing new Spotify-related markets, though a July contract on which artists would reach No. 1 in the U.S. remained open and had attracted about $894,000. Kalshi has expanded surveillance for its sports markets through a partnership with Sportradar; entertainment contracts often rely on third-party data providers that lack the same information-sharing systems.

A U.S. derivatives regulator issued an advisory in March requiring exchanges to identify the specific data sources settlements depend on and to assess those sources for reliability, objectivity and resistance to manipulation. A subsequent advisory issued this month directed exchanges to stop self-certifying broad template contracts that bundle different settlement sources under a single filing. The advisories note that cash-settled contracts can create incentives to influence the data that determine payouts.

The dispute involved settlement based on third-party data feeds that may be altered after contracts are resolved. Similar vulnerabilities have appeared previously when event data used to settle prediction markets were manipulated.