
House Oversight Committee Chair James Comer sent formal investigation letters to Polymarket CEO Shayne Coplan and Kalshi CEO Tarek Mansour on May 22, demanding documents by June 5 on identity verification, geo-restrictions, and unusual-trade detection. The letters followed a New York Times investigation that identified more than 80 Polymarket accounts placing suspiciously timed bets, including positions opened in the hours before undisclosed US and Israeli military operations against Iran. They also followed the April 24 federal indictment of US Army Master Sergeant Gannon Ken Van Dyke, accused of using classified information about Operation Absolute Resolve to net over $409,000 across prediction markets after betting on the timing of the Maduro capture.
This is not a finding. It is a probe with subpoena power behind it, aimed at the structural problem prediction markets cannot wave away. Anyone holding non-public information can monetize it directly through a binary contract that pays out the moment the event becomes public.
What Comer Is Actually Investigating
The May 22 letters from the House Oversight Committee are narrow on paper and broad in practice. Comer is asking each company for all documents and communications since January 2024 covering five buckets. Identity verification procedures and the rate at which accounts are flagged or closed. Geo-restriction enforcement and any internal data on how often US users access Polymarket through VPNs. Trade-surveillance systems, including how unusually large or suspiciously timed positions are detected and escalated. Any prior referrals to the CFTC, DOJ, or other agencies. And internal policies governing employees, contractors, or anyone with access to non-public information about world events.
The deadline is June 5. That is 14 days from the date the letters were sent, which is fast by congressional standards and signals that the committee already has a working theory rather than a fishing expedition.
The framing matters. Comer is not asking if prediction markets are legal. Kalshi is a CFTC-regulated designated contract market and Polymarket settled with the CFTC in 2022. He is asking if the surveillance and access controls these venues already have are anywhere close to what would be required if they were treated as the information-sensitive trading venues they functionally became after the 2024 D.C. Circuit ruling opened the door to event contracts on elections and political outcomes.
The 80-Account Pattern the NYT Surfaced
The New York Times investigation published five days before Comer's letters identified more than 80 Polymarket accounts whose betting patterns the paper called "statistically inconsistent with public information at the time of the trade." The largest concentration involved Iran-related contracts in March, April, and May, with several accounts opening positions on questions like "Will the US strike Iranian nuclear facilities by June 30" in windows that preceded reporting of the actual operations by 6 to 48 hours.
No individual identified in the NYT report has been charged. The accounts are linked to USDC wallets and the on-chain trail shows the wagers but not the identities behind them. That is the structural problem. Polymarket runs on an offshore, USDC-denominated order book and the entire premise of the market is that anyone in the world can take a position on any contract without revealing who they are, which works fine for entertainment betting on the Oscars and works badly for binary contracts on classified military operations.
Two patterns recur across the 80-account dataset. Accounts tend to size up sharply in the final hours before a market resolves, which is consistent with information leaks rather than analysis. And the same wallet clusters reappear across unrelated events, suggesting either a small group of insiders or a few brokers fronting trades for multiple sources.
The Van Dyke Indictment Is the Case Study
The case Comer keeps referencing in the letters is the April 24 federal indictment of Master Sergeant Gannon Ken Van Dyke, returned by a grand jury in the Eastern District of Virginia. The charges are unauthorized disclosure of classified information, wire fraud, and money laundering. The factual allegation is that Van Dyke, a signals intelligence specialist with access to operational details about Operation Absolute Resolve, opened positions across Polymarket and Kalshi on the timing of the Maduro capture and related contracts. Prosecutors say the trades cleared more than $409,000 in profit between mid-March and early April.
The Van Dyke case matters for the probe in a way the NYT pattern does not. It is a named defendant, a federal indictment, and a concrete dollar figure tied to a specific classified program. It moves the conversation from "this looks suspicious" to "this happened, here is the evidence, and the venue's surveillance did not catch it." For Comer, that is the wedge.
The indictment also surfaced an awkward detail for Kalshi specifically. Van Dyke is alleged to have used his own verified Kalshi account, with KYC completed, for a portion of the trades. A CFTC-regulated venue with full identity verification did not flag a six-figure run by a US Army master sergeant on contracts directly tied to the operations of his own intelligence specialty. That is exactly the surveillance gap the June 5 document request is built to expose.
Polymarket and Kalshi Live on Different Sides of the Jurisdiction Line
The two venues are not the same business and the probe will probably treat them differently by the time it produces findings.
Polymarket is offshore, USDC-denominated, and runs on Polygon. It controls roughly 70% of global prediction-market volume by most third-party estimates, and it geo-blocks US retail under its 2022 CFTC settlement. The block is enforced at the IP layer and is routinely bypassed via VPN, which the NYT reporting and prior CFTC settlements have already established. There is no KYC for the average user. Funding is on-chain USDC, which means every position is publicly visible on Polygon but the wallet behind it is pseudonymous. For a deeper rundown of how Polymarket actually works, the Phemex Academy explainer on how Polymarket operates covers the order book, the resolution oracle, and the offshore structure.
Kalshi is the opposite stack. It is incorporated in the US, regulated by the CFTC as a designated contract market, settles in dollars, and requires full KYC for every account. Its market share is smaller (high single digits to low teens depending on the month) but its contracts are US-legal and US-listed. The trade-off Kalshi made for that legitimacy was full identity verification, which is why the Van Dyke case is more damaging to Kalshi optically than the broader 80-account NYT pattern is to Polymarket.
|
Dimension
|
Polymarket
|
Kalshi
|
|
Jurisdiction
|
Offshore (Panama entity)
|
US, CFTC-regulated DCM
|
|
Settlement currency
|
USDC on Polygon
|
US dollars
|
|
KYC required
|
No
|
Yes
|
|
US retail access
|
Geo-blocked, VPN bypass common
|
Native US access
|
|
Global market share
|
~70%
|
High single digits
|
|
Surveillance under probe
|
Identity, geo-block, on-chain monitoring
|
KYC linkage, position-size flags
|
If you want the broader category context for how this entire venue type works, the Phemex Academy primer on prediction markets and how they price information is a good place to start before reading the June 5 disclosures.
What June 5 Will Actually Deliver
The deadline is structured to produce documents, not testimony. Comer is asking for written records and electronic communications, which is the standard opening move when the committee plans to use the production as the basis for a public hearing in late June or early July. Three things will determine if this probe ends in a quiet referral or a televised pile-on.
The first is the surveillance documentation itself. If Kalshi's records show that the Van Dyke positions were flagged internally and escalated, the company can plausibly argue the system worked and the timing was simply unlucky. If the records show the trades passed unnoticed, the CFTC will face direct pressure to mandate position-pattern surveillance comparable to what equities exchanges already run.
The second is the Polymarket geo-block data. If Polymarket's internal numbers show the VPN access rate is meaningful (one credible academic estimate is that 25 to 40% of Polymarket's traffic originates in the US through VPNs), that becomes ammunition for the argument that the 2022 CFTC settlement is functionally unenforced. The CFTC settled for $1.4 million in 2022 and Polymarket agreed to block US users. If 30%+ of usage is US-based, the settlement looks like a fiction.
The third is the referrals file. If either company has previously referred suspicious activity to the CFTC or the FBI, the committee will want to know what happened to those referrals. If neither has ever made one, despite the patterns the NYT documented, that is the headline.
Bloomberg reported on the same day the letters went out that both companies have engaged outside counsel and intend to fully comply with the document request, which is the boilerplate response and tells you nothing about what they will actually produce.
Frequently Asked Questions
Is it illegal to trade on Polymarket from the United States?
Polymarket is geo-blocked for US users under a 2022 CFTC settlement, which means accessing it through a VPN puts you outside the platform's terms of service and potentially in violation of the settlement. The bigger legal exposure is downstream. If you trade on material non-public information, the underlying conduct (insider trading, wire fraud, mishandling classified material) is illegal regardless of which venue you use.
Has anyone been charged for prediction-market insider trading before Van Dyke?
Van Dyke is the first federal indictment specifically built around prediction-market trades on classified information. There have been prior CFTC enforcement actions against operators (including the 2022 Polymarket settlement), but no individual user had been criminally charged for using a prediction market to monetize non-public information until April 2026.
Will Comer's probe shut down Polymarket or Kalshi?
A shutdown is almost certainly off the table at this stage. The probe is aimed squarely at surveillance and access-control gaps, not at the underlying legality of the venues themselves. The most likely outcomes are mandatory position-pattern monitoring, tightened KYC linkage on Kalshi, and pressure on Polymarket to actually enforce its US block. A shutdown would require either CFTC action or new legislation, neither of which is on the immediate table.
How is this different from regular financial-market insider trading?
On equities exchanges, the surveillance infrastructure is dense and the SEC has been chasing information leaks for 90 years. Prediction markets are roughly five years old as a real venue type and the surveillance stack is thin. They also pay out faster (within days or hours of an event), which makes them more attractive to anyone holding short-horizon non-public information. The economics are tighter and the controls are looser, which is the exact gap the probe is targeting.
Bottom Line
June 5 is the date that matters. If Polymarket and Kalshi produce documents showing serious internal surveillance and the Van Dyke trades were either flagged or were a genuine outlier, the probe becomes a regulatory tune-up and the venues survive intact. If the documents show what the NYT pattern and the Van Dyke indictment together imply (that six-figure positions on classified events pass through with no escalation), the next phase is a CFTC rulemaking on prediction-market trade surveillance and a hearing that puts Coplan and Mansour in front of cameras. Watch the document production volume and any pre-June 5 leaks for the early signal. The CFTC's response in the two weeks after June 5 is the lever that decides if this turns into structural rules or fades into a memo.
This article is for informational purposes only and does not constitute financial or investment advice. Cryptocurrency trading involves substantial risk. Always conduct your own research before making trading decisions.






