Get your reservation for our annual Kentucky Derby seminar now! Details Here
Why Prediction Markets and Exchange Wagering Can Never Be Fully Policed — And Why That Doesn’t Mean You Shouldn’t Swim
A man walks into a Paris weather station. He has a hairdryer and a Polymarket account. He heats the sensor. He collects $36,900. The market predicted everything. Except the hairdryer.
That story — viral, almost comedic in its audacity — is the easy version of the problem. The photogenic one. The one that gets shared on social media between racing fans and finance guys who never knew a prediction market existed six months ago.
The hard version landed last week.
On April 23, 2026, the United States Department of Justice arrested Gannon Ken Van Dyke, a U.S. Army Special Forces master sergeant, and charged him with using classified national security information to place bets on Polymarket. His alleged crime: knowing that American forces were about to capture Venezuelan President Nicolás Maduro during a covert operation called Operation Absolute Resolve — and wagering $33,000 on that outcome days before it happened. He walked away with approximately $409,000. Then he tried to delete his account.
It is, according to the DOJ, the first known criminal prosecution for insider trading on a prediction market in United States history.
It will not be the last.
The Architecture of an Unpoliced Ocean
I have been arguing for exchange wagering in horse racing since at least 2014. I want that on record, because this article is not a recantation. It is a reckoning — the kind of honest accounting that anyone who believes in these markets has an obligation to make.
Exchange wagering and prediction markets share the same foundational logic: willing participants setting prices against one another, with the platform taking a modest commission rather than a racetrack or bookmaker siphoning a ruinous takeout. The efficiency is real. The price discovery is real. The democratization of wagering is real.
So is the exposure.
When you swim in the ocean, you can get bitten by a shark. That’s true no matter how good a swimmer you are. The ocean doesn’t care about your credentials.
Prediction markets are the ocean. The sharks are not hypothetical. And the beach warnings, it turns out, have been mostly decorative.
The Van Dyke case was built on a 1980s-era commodities statute — the so-called “Eddie Murphy Rule,” Section 746 of the 2010 Dodd-Frank Act — which prohibits trading on stolen or misappropriated nonpublic government information. That law was named, with some bureaucratic whimsy, after the movie Trading Places. The CFTC invoked it for the first time against a prediction market participant last week. The legal framework was already there. It just hadn’t been used.
That tells you something important: the problem is not that the laws don’t exist. It’s that the markets grew faster than the enforcement posture. Polymarket went from a platform the CFTC fined $1.4 million in 2022 for operating without registration to a fully CFTC-registered U.S. exchange by July 2025. The legal scaffolding is still going up while the trading floor has been live for years.
A Pattern, Not an Incident
The Van Dyke arrest did not emerge from a vacuum. It crystallized a pattern that researchers, regulators, and journalists had been documenting for months.
The Iran bets. In the hours before the February 28, 2026 U.S.-Israeli strike on Iran — one of the most operationally sensitive military actions in recent memory — six newly created Polymarket wallets collectively purchased approximately $1.2 million in ‘Yes’ contracts on a U.S. strike happening by that date. One account, operating under the handle ‘Magamyman,’ placed its first trade 71 minutes before the news broke, when the market was pricing only a 17% probability. It collected roughly $553,000 when the strike was confirmed. Another Polymarket trader, operating since 2024, had accumulated nearly $1 million across a series of remarkably well-timed Iran-related positions.
The Israeli arrests. Israeli authorities have already indicted two individuals — including a military reservist — for allegedly using classified military intelligence to place bets on Polymarket during Israel’s war with Iran. These were the first publicly known arrests worldwide tied to a prediction market bet made on state secrets. They predated Van Dyke’s arrest by weeks.
The congressional candidates. Kalshi, the other major U.S. prediction market platform, this week announced it had suspended three congressional candidates from its platform for political insider trading — specifically, betting on their own electoral outcomes. Kalshi declined to refer the cases to the DOJ, and legal experts noted that betting on one’s own campaign is probably not actually illegal under current law. The absurdity of that gap speaks for itself.
The MrBeast employee. Kalshi also suspended an employee of the popular YouTube personality MrBeast who allegedly placed trades tied to what would happen on MrBeast videos before they were published. The market had become a vehicle for monetizing advance knowledge of everything from military operations to viral content.
A Columbia Law School study published in March 2026 screened over 93,000 distinct Polymarket markets and nearly 50,000 unique wallet addresses across two years. Its composite scoring system — measuring bet size, pre-event timing, directional concentration, and profitability — documented what the researchers called a systematic pattern of informed trading rather than lucky speculation. Their paper covered the range from military strikes to Taylor Swift announcements.
The title, to their credit, was blunt: “From Iran to Taylor Swift: Informed Trading in Prediction Markets.”
The Kinahan Connection: When Sharks Become Cartels
I wrote about Daniel Kinahan and his cartel’s involvement in horse racing — specifically, the extraordinary story of how a network linked to organized crime, money laundering, and multiple European murders had embedded itself in the thoroughbred industry through the world of boxing promotion and, through those connections, into the sport of kings.
That piece ran at Past The Wire. And then, on April 15, 2026 — right around the time it was circulating — Daniel Kinahan was arrested in Dubai.
He had been there for years. Openly. He had a sprawling villa, a property portfolio, a boxing promotion operation, and a UAE corporate structure. The U.S. Treasury had sanctioned him in 2022. American law enforcement had offered $5 million for information leading to his arrest. He had attended his own lavish wedding at the Burj al-Arab in 2017 with crime bosses from across the world as guests, while authorities estimated that his cartel controlled roughly one-third of Europe’s cocaine market.
He was arrested within 48 hours of Ireland issuing a warrant — after Dublin and the UAE quietly concluded an extradition treaty, ran a secret joint operation, and moved.
It took years of diplomatic architecture and one covert operation to bring in a man who had been operating in relative daylight. That is not a coincidence. That is what enforcement actually looks like in transnational crime.
The relevance to prediction markets is structural, not incidental. Kinahan’s operation was a masterclass in the exploitation of jurisdictional seams — operating from Dubai precisely because it sat outside the easy reach of European and American law enforcement. Polymarket is domiciled in Panama. Its international exchange, used by Van Dyke, operates outside the CFTC’s direct regulatory architecture. The parallel is not lost on federal prosecutors.
The CFTC itself argued successfully that Polymarket’s offshore operations still fell under U.S. jurisdiction because the company’s headquarters and staff are in New York. That is the same logic that will be used to reach actors who believe a Panama registration gives them cover. It won’t — not indefinitely.
The Honest Accounting
Here is where I am required to be candid about my own position.
I believe in exchange wagering. I have believed in it for over a decade. I believe I can make significant money in these markets — legally, without inside information, using the same analytical framework I have applied to horse racing for my entire professional life. I believe the efficiency of these markets, when properly constructed and regulated, is genuinely superior to the pari-mutuel model that has hollowed out racing handle for thirty years.
I also believe that the ocean has sharks.
The Van Dyke case is not an argument against prediction markets. It is an argument for their maturation. The SDNY’s Jay Clayton has been explicit: his office’s securities and commodities fraud unit is actively developing cases. The CFTC has filed its first civil complaint in connection with prediction market insider trading. Congress has proposed legislation — the Public Integrity in Financial Prediction Markets Act — that would prohibit federal officials and congressional staff from trading on contracts where they hold material nonpublic information.
None of these moves kill the market. They are the market growing up.
The parallel in horse racing is obvious. Racing exchange wagering has been available in the United Kingdom and Hong Kong for years. It has not eliminated cheating. It has not eliminated the manipulation of results. What it has done is create a data trail, an audit record, and a pricing mechanism that makes suspicious activity visible in ways the old fixed-odds model never could. A $2 win ticket at a racetrack leaves almost no trace. A large position on a prediction platform leaves a blockchain record.
Van Dyke tried to delete his Polymarket account on January 6 — the morning after the Maduro capture was announced. He changed his linked cryptocurrency exchange email to an account he’d created three weeks earlier under someone else’s name. He thought he was covering his tracks.
He wasn’t. The data was already there.
What Actual Enforcement Looks Like
The sophistication gap in this space is significant and real. Markets are currently priced by a mixture of informed retail participants, institutional traders, and — in a non-trivial number of cases — people with access to information that has no business touching a market.
The honest version of the enforcement picture is this: the cases that get made are the obvious ones. A soldier who bets $33,000 on a classified military operation days before it happens and then tries to delete his account is, by the standards of this industry, not subtle. The six newly created Polymarket wallets that appeared 71 minutes before the Iran strike, collectively buying a low-probability outcome that nobody else on the platform was pricing, are not subtle.
The subtle trades are not being caught. They may never be.
This is not an argument for nihilism. It is an argument for honest eyes. Every sophisticated market in history has had informed participants operating at the edge of legality or beyond it. Equity markets. Bond markets. Commodity futures. The solution was never to eliminate the market. It was to build enforcement infrastructure commensurate with the market’s size.
Prediction markets are not there yet. They are getting there. The velocity of that movement has accelerated considerably in the past six months.
Kalshi’s CEO has tried to distinguish the platform by publicly announcing suspensions — a transparency move that is partly principled and partly strategic, recognizing that self-policing credibility is a competitive asset in a regulatory environment. Polymarket’s founder Shayne Coplan has said his company went to the DOJ immediately after spotting Van Dyke’s suspicious trades. Whether these are principled stands or liability management, the effect is similar: the platforms are beginning to behave like institutions that expect to still be operating in twenty years.
That is the right direction. It is not yet sufficient.
— — —
The Takeaway for Horseplayers and Bettors
Horse racing has spent decades failing to modernize its wagering infrastructure, in part because the institutional resistance to exchange wagering in the United States has been funded, sustained, and politically organized by the same tracks and industry bodies that benefit from the current takeout regime.
The prediction market explosion is making that resistance look not just wrong, but absurd. The regulatory framework for event-contract markets now exists. The CFTC is engaged. The DOJ is engaged. Congress is engaged. The debate about whether Americans should be allowed to bet on future events is effectively over — the question now is only about the terms.
Horse racing’s window to be part of this conversation is open, but it is not infinitely open.
I have argued for years that the best version of this future involves thoroughbred racing properly integrated into the exchange wagering ecosystem — with liquidity, data transparency, and a pricing mechanism that rewards skill rather than simply taxing it. That argument has not changed. The Van Dyke case, the Iran bets, the Israeli arrests, and the hairdryer in Paris do not change it.
What they confirm is what I have said since 2014: this is an ocean. There are sharks. You swim with open eyes, you develop your edge, you understand the environment you’re operating in — and you accept that the water isn’t always clean.
The alternative is standing on the shore watching someone else swim.
RELATED COVERAGE AT PAST THE WIRE “Knowing a Horse That Won’t Win” — PTW’s original Kinahan cartel / horse racing investigation: pastthewire.com/knowing-a-horse-that-wont-win
The Sting, if you know you know:
