Trump “not happy” with prediction markets

Federal prosecutors have charged Gannon Ken Van Dyke, an active-duty U.S. Army soldier who the indictment says has served as a U.S. Army Special Forces master sergeant, with allegedly using classified information about a military operation to make more than $400,000 in prediction-market profits.

After months of online discourse around suspected insider trading on Polymarket and Kalshi, the case is a direct test for crypto prediction markets. Prosecutors allege that Van Dyke had access to nonpublic details of Operation Absolute Resolve, the U.S. operation to capture Nicolas Maduro and Cilia Flores, and used that knowledge to trade event contracts before the public announcement.

The immediate problem is whether markets built to price public expectations can remain credible when the best-informed potential trader is a person involved in the event itself.

In that setting, the price may reflect information, but the information may come from a source the market cannot fairly absorb.

Polymarket said in a reported statement that it identified a user trading on classified government information, referred the matter to the Justice Department, and cooperated with investigators.

That distinction separates Polymarket from the alleged trader while still putting the platform’s surveillance model at the center of the case.

The political overhang sharpened after Trump was asked about suspected insider trading in prediction markets tied to war and replied that “the whole world, unfortunately, has become somewhat of a casino.”

He also said he was “not happy” with prediction markets and did not like them conceptually, which stops short of endorsing the trades but does little to quiet the market’s legitimacy problem.

The suspicion driving online debate is that some successful geopolitical trades may reflect access to restricted government timing rather than better forecasting.

The Washington Post previously reported that knowledge of the Maduro operation was limited to a close circle of national security officials, and quoted Sen. Chris Murphy arguing that such bets appeared likely to come from the White House or from people with inside knowledge, while the White House denied any staff wrongdoing.

That remains inference, not evidence, tying any Trump adviser or official to the trades.

The alleged edge was classified access

The government’s theory is narrow. The alleged advantage was having advanced knowledge of operational details.

The DOJ announcement says Van Dyke was charged with unlawful use of confidential government information for personal gain, theft of nonpublic government information, commodities fraud, wire fraud, and an unlawful monetary transaction.

The agency’s release also states that the indictment is only an allegation, a caveat that should shape every reading of the case until a court ruling is issued.

The indictment alleges that Van Dyke was involved in planning and executing Operation Absolute Resolve from at least Dec. 8, 2025, through at least Jan. 5, 2026.

It also says he signed nondisclosure agreements covering classified or sensitive information connected to U.S. Army Special Operations Command work in the Western Hemisphere.

The trading timeline highlights something traders have been talking about on X for months, not just in this case, but across multiple markets. The indictment alleges that Van Dyke created a Polymarket account on or about Dec. 26, 2025, used a VPN, and bought roughly $33,934 of Yes shares across 13 Maduro- and Venezuela-related transactions between Dec. 27 and Jan. 2.

The CFTC complaint alleges that Van Dyke accumulated more than 436,000 Yes shares in the “Maduro out by January 31, 2026?” contract at an average price of about $0.074, for a cost of about $32,538.

When the contract was resolved, Yes, the complaint says he realized more than $404,000 in profit on that contract.

A low-priced event contract became a near-full-payout instrument after the public learned the outcome. The alleged edge was timing.

The CFTC says the Yes price in the January contract stayed below $0.13 from 10:00 a.m. ET on Dec. 29 through 1:15 a.m. ET on Jan. 3, except for a brief spike to about $0.22 around 10:42 p.m. ET on Jan. 2.

After President Donald Trump’s public announcement, the complaint says the price rose from $0.375 at 4:21 a.m. ET to $0.955 at 4:25 a.m. ET.

Put simply, the market repriced almost instantly once the public signal arrived. The allegation is that one trader already had access to the signal.

Why the CFTC case changes the frame

The DOJ case supplies the criminal theory. The CFTC case supplies the market theory.

That distinction is substantive because prediction markets have often been discussed as a hybrid category, part forecasting tool, part wagering interface, part crypto-native market.

The CFTC complaint treats the relevant contract as a commodity-law instrument, and the agency’s press release says this is its first insider-trading case involving event contracts.

It also says the case marks the agency’s first use of the so-called Eddie Murphy Rule to bring charges based on the misuse of government information.

The CFTC is telling prediction-market users that event contracts may sit inside its antifraud perimeter when confidential government information is allegedly misappropriated for trading.

The CFTC had already been moving in this direction before the arrest.

In a March 2026 prediction markets advisory, staff described event contracts as derivatives that can fall within swap or futures-like definitions depending on structure.

The same advisory reminded designated contract markets that they must monitor trading, prevent manipulation, and protect market participants from abusive practices.

The advisory also says misappropriation of confidential information in breach of a duty of trust and confidence can fall under CFTC antifraud rules commonly described as insider trading.

That context suggests the Van Dyke complaint fits a framework the regulator had already started building.

Polymarket’s harder test is timing

Polymarket’s reported response is important because it complicates the simplest version of the critique.

The company account is consistent across AP, ABC’s account of the case, and a WIRED report on the Polymarket trades: Polymarket said it identified a user trading on classified government information, referred the matter to DOJ, cooperated with the investigation, and said insider trading has no place on the platform.

That gives Polymarket a defensible first answer. The platform can argue that the system produced a referral rather than silence.

The harder question is what happened before the referral. If the CFTC’s allegations are right, the market transferred a large profit after the relevant information advantage had already been converted into positions.

That is the problem for the credibility of prediction markets. Detection after settlement can support enforcement, but market confidence depends on whether bad trades can be deterred, frozen, reversed, or made too risky before the payoff.

Current Polymarket US market-integrity materials describe a multi-layer monitoring program, real-time surveillance, investigations of unusual trading activity, potential sanctions, and referrals to regulators or law enforcement.

A filed Polymarket US rulebook also prohibits participants from trading on confidential information about an event outcome when doing so would breach a duty of trust or confidence.

Those materials show the direction of travel. They should be handled carefully in this case because the alleged trades occurred on Polymarket.com.

The current Polymarket US rulebook should remain in compliance context unless a source links it to the exact trades.

Still, the compliance architecture points to the likely industry answer.

Prediction markets will need restricted-person lists, better real-time anomaly detection, clearer treatment of government and contractor access, stronger settlement-source controls, and documented referral pathways.

The relevant issue has two layers. First, can a platform identify suspicious activity? Second, can the market convince ordinary users that they are trading against open information rather than insiders with advance access?

That timing question turns detection into a design problem. A market can maintain useful surveillance records while still leaving a credibility gap if the disputed trade has already settled.

The practical burden is to move suspicious-pattern review closer to trade entry, price movement, and resolution, especially for contracts tied to events known first by small groups of officials or contractors.

The next test is market design

Recent coverage of the CFTC’s prediction-market advisory framed the same pressure point before Van Dyke was named.

Event contracts have grown into a market-structure problem because their prices increasingly influence news coverage, political narratives, and investor sentiment. A distorted market can become a distorted signal.

That effect is sharper in crypto-native prediction markets because the product is built around fast settlement, public probabilities, and tradable attention.

Polymarket is a flagship example of blockchain-based event markets, using on-chain infrastructure and stablecoin settlement to turn current events into tradable Yes/No positions.

The Van Dyke case lands directly inside that model. If event markets are merely entertainment, insider-trading enforcement may look like a narrow criminal deterrent.

If event markets are becoming financial and media infrastructure, the standards are higher.

The next version of the industry has to go beyond the claim that markets aggregate information.

It must specify which information can be traded, who is restricted from trading, how suspicious activity is detected, and what happens when the outcome being traded is tied to classified government action, campaign strategy, corporate decisions, sports officiating, or any other small-group event.

There are two defensible takeaways from the case.

The first is favorable to prediction markets: the platform says it identified suspicious trading, referred the matter to DOJ, and the government brought charges. Under that reading, transparency and surveillance worked.

The second reading is more difficult for the industry: the alleged trader reportedly entered the market before public disclosure, the contract repriced after the announcement, and enforcement followed after the alleged profit was already created.

Under that view, the market can find the trail, but the market still has to prove it can protect price formation in real time.

The narrow prosecution centers on Van Dyke. The broader test belongs to prediction markets.

They now have to show that a public odds market can survive contact with private government information without turning every sensitive event contract into a bet against insiders.