Kalshi Pledges $1 Million for Research on Its Midterm Markets as Scrutiny Mounts
Key Takeaways
- •Kalshi has pledged $1 million to fund up to 50 research projects based on data from the prediction markets it hosted during the US midterm elections, with applications handled through its Kalshi Research arm.
- •Trading volume on Kalshi's 2026 election markets exceeded $2.5 billion, nearly doubling the $1.3 billion recorded during the 2024 cycle.
- •Kalshi's own research arm reported directional accuracy of 98.2% by the time polls closed, up from 93.8% three months earlier, though a RotoWire analysis found rival Polymarket held a slight edge on House and Senate odds.
- •Roughly 20 state attorneys general are challenging Kalshi, including an Arkansas lawsuit filed one day after the election and Utah felony charges against co-founders Tarek Mansour and Luana Lopes Lara.
- •The CFTC has privately warned Kalshi it could face charges over its self-certification of dozens of contracts, and a settlement could force slower contract-by-contract filings on which much of the sector's growth depends.

Kalshi has pledged $1 million and data from the prediction markets it hosted during the US midterm elections to fund up to 50 research projects producing studies based on the dataset. Prediction markets are exchanges where contracts on real-world outcomes trade at prices that function as probability estimates.
The offer from the leading prediction market platform came shortly after CEO Tarek Mansour publicly celebrated the accuracy of the odds it supplied as America went to the polls. It also lands amid fresh uncertainty: the elections produced a Democrat-led Congress, while renewed scrutiny from the Commodity Futures Trading Commission (CFTC), the federal agency that oversees US derivatives markets, is adding to pressure from the roughly 20 state attorneys general who have the company in their sights.
A $1 million grant program open to researchers
Under the program, interested researchers can submit pitches for consideration by quote-tweeting their ideas to the post announcing the initiative and applying through the company's website. Successful applicants will have access to any public platform data they need for their research projects.
The program is run by Kalshi Research, a unit launched almost one year ago, in December 2025, to build a route connecting academics with the platform's internal trading records and to host a prediction market conference. Kalshi has compared the research arm to similar units at leading AI labs such as Anthropic and OpenAI. At launch, the platform said it had onboarded scholars from schools including Harvard, Stanford, the University of Chicago and Yale.
Data from the research arm over the past year reflect the scale of the platform's growth. One research paper reviewing the 2026 midterms reported directional accuracy reaching 98.2% by the time polls closed, improving on its 93.8% score from three months earlier. Brier scores, a measure of forecast error in which lower is better, moved from 0.040 to 0.016 over the same period. The study said Kalshi matched or beat polling averages at every point where polls existed. Those figures come from Kalshi's own research arm; the new grants fund outside projects built on the same data.
Strong results from the November midterms
Kalshi put up strong numbers during the November midterms, with trading volume topping $2.5 billion on its 2026 election markets — nearly double the $1.3 billion from the 2024 round.
Mansour took a victory lap on X, declaring that "The power to decide what America sees changed hands," and pointing to the platform's record of calling winners in about 98% of the races it listed. Co-founder Luana Lopes Lara dubbed Kalshi "the future of news." The platform reported that listed favorites emerged as winners in 479 of the first 487 decided contests.
Notably, rival prediction market platform Polymarket held a slight edge in accuracy over Kalshi's House and Senate odds, according to a RotoWire breakdown.
Democratic Congress, state lawsuits and potential CFTC action
As Kalshi and Polymarket rightly forecast, Democrats took both the House and Senate on November 3, and several of the party's lawmakers are not big fans of prediction markets as they currently operate. Senator Jeff Merkley and Representative Maxine Waters have criticized the platforms for letting people bet on elections, war and terrorism. CNBC has counted nearly 30 prediction market bills introduced since 2025, most from Democrats, according to Cryptopolitan's reporting.
The pressure is not strictly partisan. Senator Elizabeth Warren, a Democrat, and Senator John Curtis, a Republican, sent the platforms joint letters before the vote asking how they guard against insider trading.
State courts are already a battlefield. Arkansas Attorney General Tim Griffin sued Kalshi on November 4, one day after the election, calling it a "pocket-sized bookie" operating without a gaming license. That suit pushed the number of states challenging Kalshi to roughly 20, with Utah charging co-founders Mansour and Lopes Lara with 13 felonies in late October.
Then there is the regulator Kalshi has leaned on for cover. Politico reported on November 10 that the CFTC privately warned Kalshi it could face charges over how it self-certified dozens of new contracts — a fast-track filing process that lets exchanges list contracts on their own certification unless the agency objects, and one the CFTC has flagged twice this year, as Cryptopolitan previously reported. Kalshi spokesperson Elisabeth Diana described the matter as a procedural error involving a handful of contracts, many of which never went live.
A settlement there could force exchanges into slower, contract-by-contract filings — the speed on which much of the sector's growth depends. How the CFTC question resolves, how the roughly 20 state challenges play out, and what the first grant-backed studies find in the midterm data are the threads to watch from here.
Source: Cryptopolitan