Prediction markets are fueling a new era of political graft

In the days leading up to the 2026 State of the Union, George Santos was acting—at least in public—like a man preparing to take a victory lap. In October 2025, President Donald Trump had commuted the former Republican congressman’s 87-month federal prison sentence, which he’d received in April 2025 after pleading guilty to various forms of campaign funds-adjacent fraud. Four short months later, on X, Santos began teasing his triumphant return to the chamber from which he’d been expelled two years earlier. “I’m going to be there for the State of the Union, in the gallery, guys,” Santos said in a video posted the night before the main event. “Just chill, trolls. Chill.” In other tweets, he floated the idea of wearing a “bedazzled” suit to the event, before eventually telling his followers that he’d opted for a “basic” blue or gray number instead. On Kalshi, a prediction market where users were betting on (among many other things) whether Santos (among many other people) would attend the State of the Union, his announcement caused the price of YES shares to spike. Then came the plot twist: At about 6 PM on Tuesday, Santos tweeted, “Watching SOTU from an airport tv was not part of the plan!” He added two red-faced emojis, and a rueful three-letter acronym expressing frustration that even the most aggressively online millennials stopped using a decade ago. Almost immediately, the price of YES bets on Santo’s attendance tumbled into single-digit territory. Sure enough, when Trump began speaking, Santos was nowhere in sight. A major snowstorm earlier in the week had snarled air travel up and down the East Coast, and on Wednesday, when a Kalshi user complained on X that they’d lost money because of him, Santos blamed a problem with a flight. “Trust me I wanted to be there too,” he wrote. As it turns out, Santos may have found a way to make the best of a bad situation. According to NPR, before he sent that market-flipping tweet, Santos hopped on Kalshi to bet that he would not be at the State of the Union after all, which earned him a five-figure payout. As a result, NPR says, Santos is now the subject of investigations by the Commodity Futures Trading Commission, the federal agency that regulates Kalshi, as well as the U.S. Department of Justice, which raises the hilarious possibility that Santos, now eight-ish months out of prison, could find himself back in custody in the not-so-distant future. Among the oldest, grimiest traditions in Washington is its revolving-door culture, which enables outgoing politicians and former staffers to leverage their experience by taking well-compensated positions within the same industries they used to regulate. But Santos’s latest alleged ethical misadventure is a newer form of corruption, made possible by the explosive growth of bet-on-everything platforms that the Trump administration has zero interest in meaningfully checking. When trading on access is this frictionless, enterprising political types with inside information are going to look for ways to turn a profit. This information arbitrage can be very lucrative. In January 2025, a bettor on Polymarket, a Kalshi competitor, netted more than 300,000 by placing long-odds bets that President Joe Biden would pardon four specific individuals during his final hours in office. In January 2026, another Polymarket user made more than 400,000 on a 32,000 bet that the United States would capture Venezuelan president Nicolas Maduro by the end of the month. On February 28, 2026, six Polymarket wallets collectively earned more than a half-million dollars by correctly predicting that the United States military would strike Iran by the end of that day. Prediction markets know that stories like these are (to use a technical term) not a good look. In April, Kalshi suspended the accounts of three candidates who’d wagered on the outcomes of their races, including one who “predicted” that he would enter the race two months before he did so. But there is just too much money flowing here to keep all the insiders out. In May, NPR spoke with campaign staffers who were making money after seeing not-yet-released information that they knew would scramble the prediction markets. One staffer described taking positions on Kalshi based on a not-yet-released poll that showed their candidate in an unexpectedly commanding lead. “Because you have all this information and knowledge that isn’t publicly available yet, it’s almost foolish not to bet on it before it’s made public,” they said. Two years ago, when Kalshi began rolling out its politics markets to coincide with the 2024 election, I worried that allowing people to invest both emotionally and financially in their preferred candidates was one of the few remaining methods of making politics in this country worse. In the weeks before Election Day, Trump had supplemented his usual voter-fraud conspiracy theories with warnings that he would only lose the election if his enemies rigged it against him. Given that Trump supporters responded to his loss in 2020 by storming the Capitol, I don’t imagine that they would have reacted more reasonably if they had also gambled away their life savings in the process. What stunts like Santos’s show, though, is that the harms of political gambling include well-timed rug-pulls, in addition to the tawdry reduction of democracy to a yes-or-no bet on whether the Supreme Court will allow a xenophobic president to revoke the Fourteenth Amendment’s guarantee of birthright citizenship by executive fiat. As the Polymarket incidents suggest, people are already making big money betting on things they have good reason to believe will happen. Santos went one step further: He knew his bet was a sure thing because he could exercise some control over the outcome. Relative to the deposition of a foreign head of state or the initiation of a stupid and illegal war, a disgraced former lawmaker’s presence at a State of the Union is pretty inconsequential. But it is not hard to imagine savvy bettors using the same basic playbook when the stakes are higher. There are markets, for example, about if and when particular politicians will leave office, and how they will vote on specific matters. Once they make up their minds, it would not be especially complicated for them (or someone close to them) to make some money off all the rubes out there who are just taking an educated guess. In theory, this sort of insider trading is illegal. In practice, under Trump, the CFTC has been far less interested in scrutinizing prediction markets than in protecting them from anyone who would dare to try. Chairman Michael Selig has praised prediction markets as “innovative financial products,” and warned of the dangers of attempting to “regulate them away.” Shortly after Minnesota Democratic Governor Tim Walz signed a law banning prediction markets, Trump called Walz “SCUM” for challenging the CFTC’s “exclusive authority.” The three candidates who wagered on their races did get hit with modest fines and five-year trading bans. But as PBS notes, these were agreements between the candidates and Kalshi, not punishments imposed by the CFTC or other law enforcement agency, and the CFTC only found out about Santos because Kalshi reported him to it. As ever, allowing industries to police themselves is tantamount to allowing industries to do as they please. It is of course not new for people who work in politics to have access to information that is, in the general sense, valuable. But what platforms like Kalshi and Polymarket have done is commodify that information to a greater extent than ever before, allowing anyone with smartphone access to profit off exactly what they know in just a few taps. If you have been thinking about dabbling in insider trading, now is the time to try, because the chances are pretty good that the powers that be won’t find out. They probably aren’t looking in the first place.