Luck alone cannot explain those numbers.
In the intersection of financial speculation and national security, a new form of insider trading has emerged: military personnel and connected insiders are allegedly using classified knowledge of U.S. operations to place bets on prediction markets, turning state secrets into private fortunes. An Army master sergeant's arrest for wagering on a raid he helped plan is not an isolated breach but a visible edge of a far wider pattern, one in which the mathematics of improbable wins—98% success rates, $2.4 million in coordinated profits—speak louder than any single indictment. The same markets designed to aggregate collective wisdom are now functioning as mirrors of classified American intent, readable not only by investigators but potentially by adversaries. What began as a question of financial ethics has become a question of whether the architecture of modern speculation is compatible with the secrecy on which military operations depend.
- A soldier who helped plan a covert raid on Venezuela's president placed $34,000 in bets the day before it happened and walked away with over $400,000—then tried to delete the evidence.
- Researchers found nine linked accounts with a 98% win rate on U.S. military bets, collectively earning $2.4 million, while correctly predicting the precise dates of strikes and ceasefires in the Iran war against overwhelming odds.
- Over $800 million was suddenly wagered on falling oil prices just fifteen minutes before President Trump announced ceasefire talks with Iran, suggesting classified diplomatic knowledge may be flowing into commodity markets as well.
- A journalist who published a story that voided military-outcome bets received death threats backed by $22 million in wagers, with senders demonstrating knowledge of his family members' daily movements.
- The federal agency responsible for oversight has seen enforcement actions drop by more than two-thirds since 2024, leaving a widening gap between the scale of the problem and the capacity to address it.
- National security officials warn that the same irregular betting patterns visible to investigators are equally visible to foreign adversaries, who could use them to anticipate and counter U.S. military operations in real time.
An Army master sergeant with access to classified military plans placed six bets on Polymarket the day before a covert operation captured Venezuela's Nicolás Maduro. He collected more than $400,000, then attempted to delete his account. Charged in April by the Justice Department, he has pleaded not guilty—but investigators say his case is not an anomaly. It is a symptom.
This year, more than a billion dollars has been wagered online on military decisions and outcomes. The mathematics of these bets are damning. Researchers at the Anti-Corruption Data Collective found that long-shot military bets on Polymarket succeeded more than half the time—a 52% win rate—compared to just 7% for sports bets. A Paris-based analytics firm, Bubblemaps, identified nine linked accounts that earned $2.4 million betting almost exclusively on U.S. military operations, with a 98% win rate. They correctly predicted the dates of the first U.S. strikes on Iran, the removal of Iran's supreme leader, and the ceasefire announcement. "This is like winning the lottery multiple times," one investigator said. "Luck alone cannot explain those numbers."
The corruption reaches into commodity markets as well. On March 23rd, more than $800 million was suddenly staked on falling oil prices. Fifteen minutes later, President Trump announced productive ceasefire talks with Iran. Oil dropped more than 10 percent. The potential profit: $80 million. Federal investigators are now probing those trades alongside the prediction market cases.
The phenomenon has also turned dangerous for those who report on it. When journalist Emanuel Fabian published a brief article about an Iranian missile landing in an empty forest near Jerusalem, he received threats from bettors who stood to lose $900,000. His story had voided bets predicting no missile strike that day—part of $22 million wagered on the outcome. The threats included details about his siblings' movements. Polymarket banned the accounts involved, but the episode revealed something troubling: when enough money is at stake, people will attempt to rewrite the historical record through intimidation.
The Commodity Futures Trading Commission, the agency responsible for policing these markets, has seen its enforcement actions drop by more than two-thirds since 2024. Its staff has shrunk. It is now led by a single 36-year-old chairman nominated last fall, who says the agency is deploying artificial intelligence to catch bad actors. A second case has since emerged—a Google software engineer allegedly used inside information to earn over a million dollars on Polymarket. A Defense Department source described Van Dyke as "just a small fish."
What most alarms national security officials is not the financial theft but the intelligence leak it represents. Irregular betting patterns are visible to anyone watching the markets—including foreign adversaries who could read them to anticipate U.S. military movements and adjust their own strategies accordingly. The prediction market, conceived as a tool for aggregating dispersed knowledge, has become an open window into classified American planning.
A soldier with access to classified military plans placed $34,000 in bets the day before a covert raid on Venezuela's president. When the operation unfolded exactly as he had wagered, he collected more than $400,000. Then he tried to erase the evidence. This is no longer an abstract concern about market manipulation. It is happening now, in real time, on platforms designed to let anyone anywhere bet on anything—including the movements of U.S. military forces and the fates of world leaders.
Army Master Sergeant Gannon Ken Van Dyke was charged in April with using classified intelligence to trade on Polymarket, the world's largest online prediction market. According to the Justice Department, he knew the exact timing of the January 3rd operation that captured Nicolás Maduro because he helped plan and execute it. He placed six bets the day before the raid unfolded. The indictment alleges he immediately withdrew his winnings and attempted to delete his account. Van Dyke has pleaded not guilty. But the case is not an isolated incident. It is, according to investigators and market analysts, a visible symptom of something far larger.
This year alone, more than a billion dollars has been wagered online on military decisions and outcomes. Bettors across the globe have taken positions on when attacks might occur, how they might unfold, even the survival of foreign leaders. Some bets are placed with information so specific, so precisely timed, that they could only come from someone with access to classified planning. Rob Schwartz, a former official at the Commodity Futures Trading Commission, calls this a new kind of insider trading. "If you are a corporate executive privy to non-public business information and you trade on that, everybody knows that's insider trading," he said. "The same thing exists in prediction markets. This is just a different context."
What makes the pattern unmistakable is the mathematics. Michelle Kendler-Kretsch and her team at the Anti-Corruption Data Collective examined long-shot military bets on Polymarket—wagers with less than 35 percent odds of winning. Despite being underdogs, they succeeded more than half the time. Military bets showed a 52 percent success rate. Sports bets, by contrast, succeeded 7 percent of the time. The disparity is not subtle. It is the signature of systemic insider trading. A Paris-based data analytics firm called Bubblemaps went further. Researchers there identified nine linked accounts that collectively earned $2.4 million betting almost exclusively on U.S. military operations. Their win rate was 98 percent. They correctly predicted the dates of pivotal moments in the war with Iran—the first U.S. strikes, the removal of Iran's supreme leader, the announcement of a ceasefire—even when the odds were heavily stacked against them. "This is like winning the lottery multiple times," one investigator said. "Luck alone cannot explain those numbers."
The scope of the problem extends beyond prediction markets into traditional commodities trading. On March 23rd, as fighting with Iran had raged for more than three weeks, oil futures were trading slowly. Then, at 6:50 a.m., more than $800 million was suddenly staked on oil prices dropping. Fifteen minutes later, President Trump posted on Truth Social that the White House and Iran had reached "very good and productive conversations" about ending hostilities. Oil prices plummeted more than 10 percent. The profit from that single trade could have reached $80 million. Federal investigators are now probing these oil market transactions as well.
The corruption extends into the realm of information itself. Emanuel Fabian, a military correspondent for the Times of Israel, published a brief article in March about an Iranian missile strike in an empty forest near Jerusalem. Soon after, he received messages demanding he change his story. One threat read: "You're gonna make us lose $900,000. And we'll invest even more than that to finish you." The sender included details about Fabian's siblings and their movements. He investigated and discovered that $22 million had been wagered on Polymarket on whether an Iranian missile would enter Israel on March 10th. His article had voided the bets predicting no missile strike, enraging the losers. Fabian reported the threats to police and to Polymarket, which banned the accounts involved. But the incident exposed a darker dimension of the betting phenomenon: when enough money is at stake, people will threaten journalists to alter the historical record.
The federal agency responsible for policing this landscape is the Commodity Futures Trading Commission, established in the 1970s to regulate food prices. Today it is led by a single chairman, 36-year-old Michael Selig, nominated by President Trump last fall. But enforcement actions have dropped by more than two-thirds since 2024. Staffing has declined sharply. The CFTC says it is hiring and deploying artificial intelligence to catch bad actors, but the gap between the scale of the problem and the resources devoted to solving it is widening. A Defense Department source told investigators that Van Dyke was "just a small fish." Since the initial investigation aired in May, federal prosecutors have brought another case—this time against a Google software engineer who allegedly used inside information to place bets on Polymarket and walked away with more than a million dollars in profit.
What troubles national security officials most is not the theft itself but the signal it sends. If market watchers can spot irregular trades, so can adversaries. They can read the betting patterns, identify the anomalies, and infer what the U.S. military is about to do. They can adjust their own strategy accordingly. "This could be putting people's lives at risk," one former military officer said. "Other adversaries may be using this information in order to plan their own strategy." The prediction market, designed as a tool for aggregating dispersed knowledge, has become a window into classified American military planning—one that is open to anyone with money and access to the right information.
Notable Quotes
If you are a corporate executive privy to non-public business information and you trade on that, everybody knows that's insider trading. The same thing exists in prediction markets.— Rob Schwartz, former CFTC official
This could be putting people's lives at risk. Other adversaries may be using this information in order to plan their own strategy.— Former military officer and investigator
The Hearth Conversation Another angle on the story
How does a soldier even get access to a platform like Polymarket if it's illegal for U.S. military personnel to use it?
The platform itself is based overseas and operates in a legal gray area. There are digital workarounds—VPNs, cryptocurrency transfers, accounts registered under false names. The barrier to entry is low if you know what you're doing.
But $400,000 in profit from $34,000 in bets—that's a 12-to-1 return. How does that not immediately trigger alarms?
It should. And in Van Dyke's case, it did eventually. But the sheer volume of trades happening across these platforms makes detection difficult. Polymarket processes millions of bets. Without specific intelligence pointing to a particular account, the suspicious trades can hide in plain sight.
The 98 percent win rate on those nine linked accounts—is there any scenario where that's legitimate?
Theoretically, yes. Someone could be an exceptional analyst with deep expertise in military affairs. But the odds of that happening across dozens of bets, on specific dates, during an active war? The probability is so vanishingly small that it crosses into statistical impossibility. That's what makes it evidence.
What worries you most about this—the theft itself, or what it signals?
The theft is bad. But the real danger is that adversaries are watching the same markets we are. If they see a spike in bets predicting a U.S. strike on a particular date, they can move their assets, evacuate personnel, change their plans. We're essentially broadcasting our military intentions to our enemies through the betting patterns of insiders.
Why hasn't Congress moved to shut down military betting on these platforms?
That's the question everyone's asking. The CFTC is understaffed and underfunded. The platforms operate internationally, which complicates jurisdiction. And there's a philosophical question about whether you can even regulate something that's fundamentally about information—about who knows what, and when.