Prediction Markets Are Crossing a Dangerous Line
On April 23, federal prosecutors quietly crossed a legal threshold that has been looming for years. For the first time in U.S. history, a criminal case was filed alleging insider trading—not in stocks, not in options, but in a prediction market.
The defendant was not a hedge fund manager or a corporate executive. He is a U.S. Army Master Sergeant with a top-secret clearance. According to prosecutors, Gannon Ken Van Dyke used classified information about a covert operation targeting Venezuelan President Nicolás Maduro to place roughly $33,000 in bets on Polymarket, a crypto-based prediction platform. The wagers paid out more than $409,000.
In this case, a government insider allegedly turned national security intelligence into a personal trading strategy.
That case would be significant on its own. But just days later, the two largest prediction-market platforms, Kalshi and Polymarket, are rolling out a new product: leveraged “perpetual futures,” allowing users to place bets up to ten times their account balance on real-world events, with no expiration date.
This first insider-trading prosecution has arrived at the exact moment these markets are becoming more powerful, more accessible, and potentially more dangerous.
What Are Prediction Markets
Prediction markets are often described as a hybrid between financial exchanges and sports betting apps. At their core, they allow users to buy and sell contracts tied to the outcome of real-world events, such as ‘will a political candidate win an election’ or ‘will Federal Reserve cut interest rates.’
Each question becomes a tradable contract, typically priced between a penny and a dollar, representing the perceived probability of an outcome. If the event happens, the contract pays out $1. If not, it goes to zero.
The appeal is intuitive. The interface feels like FanDuel or DraftKings, in its ease of navigate and understand. The stakes feel familiar. But the underlying mechanics are closer to derivatives trading than gambling.
Two companies dominate the space, and they operate in fundamentally different ways.
Kalshi is a U.S.-based exchange regulated by the Commodity Futures Trading Commission (CFTC). Users sign up with verified identities and connect bank accounts. It operates within a formal legal framework, positioning itself as a legitimate financial market.
Polymarket, by contrast, emerged offshore and runs on blockchain infrastructure. It accepts cryptocurrency and historically restricted U.S. users, though in practice Americans have continued to access it through crypto wallets and VPNs. Its structure allows for anonymity and global participation, with fewer traditional compliance barriers.
Together, these platforms now process billions in monthly trading volume.
Insider Trading Problem No One Fully Understood
Prediction markets have long been promoted as tools for aggregating information. The idea is simple: if enough people with diverse knowledge participate the market price becomes an accurate forecast.
In traditional financial markets, insider trading laws prohibit individuals from using material nonpublic information to profit from securities trades. In sports betting, leagues and regulators impose strict prohibitions, players, coaches, and officials are barred from betting on games they influence or have inside knowledge about.
Prediction Markets allow anyone to bet on real-world events. But unlike sports, there is no clear boundary separating participants from insiders. And unlike stocks, the scope of “material nonpublic information” extends far beyond corporate boardrooms.
The Van Dyke case illustrates the extreme end of that risk. A military insider allegedly used classified intelligence about a covert operation to place trades.
But the broader exposure is much wider, and far less understood. A pharmaceutical employee who knows the outcome of a pending drug approval or a congressional staffer aware of a committee vote before it becomes public.
If a prediction market offers a contract tied to any of those events, the legal framework begins to resemble insider trading in financial markets.
A Market That Has Already Been Tested
The Van Dyke case is not emerging in isolation. Over the past several months, a series of events has highlighted how prediction markets can be influenced—or exploited—by those with privileged information.
In early 2026, roughly $500 million in trading volume moved through Polymarket contracts tied to U.S. and allied military operations involving Iran. Blockchain analysis identified clusters of wallets placing highly accurate bets shortly before key developments. In one instance, a group of newly created accounts achieved near-perfect outcomes around a military strike.
Separately, Israeli authorities arrested soldiers accused of using classified intelligence to trade on similar markets during a regional conflict.
Regulators in the United States have begun to respond. In February, the CFTC issued its first formal advisory warning about insider trading risks in prediction markets. The agency cited cases involving a political candidate betting on his own race and a media professional trading on unreleased content.
By March, both Kalshi and Polymarket updated their rules to explicitly prohibit trading based on confidential information, a coordinated move that underscored growing concern. In April, Kalshi suspended multiple congressional candidates for betting on their own elections, including one who publicly acknowledged doing so to challenge the regulatory framework.
The Leverage Inflection Point
These products allow users to amplify their exposure. A trader with $1,000 in an account can take a $10,000 position on whether an event occurs. Unlike traditional contracts, perpetual futures do not expire, they remain open indefinitely, subject to funding rates and margin requirements.
In financial markets, leverage is a powerful tool and a well-known source of risk. It magnifies gains, but also losses. It can accelerate market movements and amplify volatility. In prediction markets, leverage introduces an additional dimension: it increases the potential payoff for anyone with privileged information. If a user already has an informational edge, leverage allows them to scale that advantage dramatically.
This is where the regulatory tension becomes more visible. The same agency that issued its first insider-trading warning on prediction markets earlier this year has allowed the rollout of these leveraged products.
Prison For Military Personnel
If convicted in federal court, Van Dyke would almost certainly serve his sentence in the Federal Bureau of Prisons system, not a military brig, which is located at FCI Leavenworth (Kansas). Because prosecutors charged him as a civilian defendant rather than through a court-martial, his case falls under the same framework as any federal financial or national security crime. His eventual placement would depend on factors like sentence length, security classification, and whether the government views him as requiring protective custody due to his background and cooperation risk. In practice, that could mean a low- or medium-security federal facility, or even a more controlled environment if officials determine his prior military role and the nature of the offense create safety concerns. The key point is that by bringing the case in civilian court, the government is treating the alleged conduct not as a breach of military discipline, but as a federal crime with consequences that extend well beyond the armed forces. The core challenge facing regulators is not simply whether prediction markets should exist. It is whether they can be effectively governed.
A Little History Lesson
For historical context, the closest analogue isn’t a single named “operation,” but the wave of insider-trading prosecutions that grew out of overlapping federal investigations in the late 2000s and early 2010s, most notably the Galleon case and what prosecutors sometimes described as a broader crackdown on “expert network” abuses. Those cases established a key principle that applies directly here: trading on material, nonpublic information whether it comes from a corporate insider, a consultant, or a government source can trigger criminal liability even if the “market” isn’t a traditional stock exchange. The Van Dyke case effectively extends that logic into prediction markets, where the underlying asset isn’t a company but an event. What’s new is the venue, not the theory. The government is signaling that informational advantage, when it crosses into misuse of confidential or classified data, is still insider trading regardless of where the bet is placed.
Tom Hardin, the former hedge fund analyst who cooperated with the FBI in the Galleon investigation as the informant “Tipper X” and author of the national bestseller Wired on Wall Street, was himself convicted of insider trading and has spent the years since speaking publicly about how people cross the line. He sees Van Dyke as an important marker. “By March, you could see hundreds of millions moving in prediction markets around wars, military strikes, and presidential actions, and the patterns were obvious to anyone watching,” Hardin says. “But this is not just a financial integrity problem. When money floods into a contract on whether the U.S. will strike a foreign target, the price itself becomes a public signal. A foreign intelligence officer does not need to hack the Pentagon. He just needs to watch the order book.”
Hardin notes that Van Dyke’s exposure is being driven by classified information statutes, which carry far steeper penalties than ordinary insider trading law. What concerns him more broadly is the psychology these platforms invite. “Stocks come with the understanding that insider trading is a crime. Prediction markets feel like entertainment, the same mental bucket as FanDuel or DraftKings. The moral structure is identical, but the language is softer, and that is exactly how people give themselves permission. If you know tomorrow’s answer today because of your office, your duty, or your security clearance, and you profit before the public can, it is insider trading whether it happens through a hedge fund trade or thirty seconds on an app.,” Hardin told me in an interview.
Bigger Picture
Prediction markets were originally framed as tools for forecasting, a way to harness collective intelligence and produce more accurate predictions about the future.
That vision still holds appeal. In certain contexts, these markets have demonstrated an ability to aggregate information effectively. But the recent wave of events highlights a different reality. When markets allow trading on real-world events, and when participants include individuals with privileged or confidential information, the line between prediction and exploitation becomes harder to draw.
The Van Dyke case did not create that tension. It exposed it. And as leverage increases, participation expands, and regulatory frameworks evolve, that tension is likely to become more visible, not less.
The question now is not whether prediction markets will continue to grow. It is whether the systems designed to govern them can keep up. Because if they cannot, the next case may not involve $409,000. It may involve much more, and far more people who never realized they were taking that risk.
Read the full article here