In the early hours of 13 June, more than 200 Israeli fighter jets began pummeling Iran with bombs, lighting up the Tehran skyline and initiating a 12-day war that would leave hundreds dead.
But for one user of the prediction market Polymarket, it was their lucky day. In the 24 hours before the strike occurred, they had bet tens of thousands of dollars on “yes” on the market “Israel military action against Iran by Friday?” when the prospect still seemed unlikely and odds were hovering at about 10%. After the strike, Polymarket declared that military action had been taken, and paid the user $128,000 for their lucky wager.
But was it just luck?
Polymarket is an online platform where people can bet on just about anything, from what the most-streamed song on Spotify will be to how many times Donald Trump will say “terrible” that day. Under Joe Biden, the unregulated site was banned in the United States, although it could be accessed with a VPN. Now, over $100m worth of bets are placed on Polymarket each day.
The user who had placed the bet was new to Polymarket. Since then they have lodged five bets on similar markets like, “Israel announces end of military operations against Iran before July?” and “Israel strikes Iran by January 31, 2026?” All have been successful, making further profits of more than $28,000.
The Guardian’s analysis of blockchain data traced the wallet to an X account with its location set to Beit Ha’shita, a kibbutz in northern Israel. The user recently transferred their bets to two other accounts, data shows, potentially to avoid detection. There, they currently have 10 live bets relating to Israel’s military strategy.
The X user behind the account could not be reached for comment but Israel Defense Forces and the Israel Security Agency have opened an investigation into the user, officials said last Thursday.
Unlike sports betting sites like DraftKings, where users have to scan their IDs before they are able to place a bet, Polymarket is built on a blockchain. Accounts are linked to crypto wallets that can be publicly traced but are difficult to link back to an individual. So, even though the company is headquartered in the US, its markets are global and the users anonymous.
Online prediction markets found mainstream traction in 2024 by letting everyday Americans wager on the US presidential election.
Americans were already becoming hooked on online gambling after the 2018 supreme court ruling that legalized sports betting. But that election supercharged the desire to bet: the twists and turns of the presidential race meant it already felt like a spectator sport.
Polymarket’s virality was initially powered by manosphere influencers. The Barstool Sports founder Dave Portnoy broadcast election odds to more than 200 million followers, right-leaning X accounts weaponized the markets, which were predicting a Trump victory, against mainstream polls, which showed a tight race. Crypto streamers pitched political betting as just another high-risk, high-reward trade.
Since then, Polymarket and its rival Kalshi have gamified all corners of life – from tomorrow’s weather to Bad Bunny’s Super Bowl outfit – and in effect, turned reality itself into a casino.
To bet on an outcome, users buy either a “yes” or “no” event contract. These are priced between $0 and $1, and like any market, their price fluctuates depending on demand. If a “yes” contract is worth $0.80, the market thinks it’s 80% likely. If the “yes” thing happens – a team wins their game, a director wins their Oscar, a bomb drops on a city – the value of the correct contract shoots up to $1, while the wrong contract drops to $0. If you buy thousands of contracts for something that seems really unlikely, $0.10 for Israel striking Iran, say, then when that contract reaches $1 – you stand to make a huge profit.
For example, one account created in November has been solely dedicated to wagering that Trump will appoint Kevin Warsh as the next chair of the Federal Reserve, persisting even when the odds dropped to 7%. After Trump announced Warsh this morning, they will profit $142,000.
Polymarket disciples reject the idea that they are merely gambling, they see the markets that result from the bets placed on the site as critical sources of information. The thinking is that Polymarket odds, based on the guesses of people willing to put their money where their mouth is, are a better method of predicting the future than flawed polls or biased commentators. They see Polymarket, primarily, as journalism.
“It’s the most accurate thing we have as mankind right now, until someone else creates some sort of super crystal ball,” Polymarket’s CEO, Shayne Coplan, told 60 Minutes in November, shortly after the 28-year-old was anointed the world’s youngest self-made billionaire, according to the Bloomberg Billionaires Index.
Polymarket provides a “credible, lightning-fast news channel for hundreds of millions of people”, said the blockchain analyst Andrew 10 Gwei [he goes by his X username], who has been exposing insider activity within crypto startups since 2020 and has more recently turned his focus to Polymarket. “You gain access to critical information about key world events faster than everyone else.”
Polymarket called the 2024 presidential race before the polls. The 2026 Golden Globes announced real-time Polymarket odds as the telecast headed into commercials, correctly predicting 26 out of 28 winners and taking a lot of the suspense out of the ceremony itself. Media entities such as CNN and the Wall Street Journal have partnered with prediction markets and integrated their data into their reporting.
But this is not the first time America has used prediction markets to hedge outcomes.
A century ago, political betting was rampant on Wall Street. Bookmakers correctly predicted 11 out of 15 elections between 1884 and 1940. These odds would be splashed across newspapers each election cycle – “never wrong”, as some papers put it. During some election years, the betting on political outcomes would exceed trading in stocks and bonds.
However, election betting wouldn’t survive the Prohibition era’s moral clampdown on gambling. Critics warned these markets risked election tampering and information withholding. At the same time, the arrival of scientific polling provided an uncontroversial alternative to finding out the mood of the nation. After the success of Gallup in predicting the 1936 election, many newspapers scrubbed Wall Street odds from their pages.
But today’s gambling revival looks nothing like the past: betting is now borderless, anonymous and limited only by imagination.
“We’re still trying to understand how these markets are going to affect our lives, but honestly, I think they’re affecting us more than we probably realize,” said Yesha Yadav, a professor at Vanderbilt law school. When Yadav asked her students recently if they bet on prediction markets, a sea of hands shot up. “It’s just becoming normal,” she said.
Event contracts are considered derivatives, so online prediction markets fall under the jurisdiction of the US Commodity Futures Trading Commission (CFTC), who, under Biden, had fought hard to block them.
Election markets are “contrary to the public interest”, Rostin Behnam, former CFTC chair said in 2023, adding that regulating them would be both “impractical” and force the agency to act as “an election cop”.
But under Trump, the CFTC has pulled back. It dropped its legal fight to ban Kalshi’s election markets, and permitted Polymarket to operate domestically under its jurisdiction.
Business has been booming ever since: the total value of bets bought on the sites in just the month of December surpassed $8.3bn, up 1,300% from February, according to DefiLlama data. That’s about the same as the monthly sales of Target.
Yet, as the list of markets grows longer, so does the number of people sitting on valuable information. Concerns have been mounting about alleged insider trading.
The world first caught on to the problem after the US capture of Venezuela’s leader, Nicolás Maduro, in January. Shortly after, reports circulated that a new Polymarket account had profited more than $400,000 by betting Maduro would be ousted by the end of January. The wagers were placed in the days and hours running up to Trump revealing the capture on Truth Social.
Yet, the profiteering may run deeper than this. The Guardian identified more than a dozen Polymarket accounts with the hallmarks of possible insider trading: large, accurate bets lodged shortly before the event, with little other activity – and are not hedging their bets across multiple accounts. Collectively, the accounts have profited about $2.17m.
The bets that weren’t gambles
The markets with suspicious activity cross many areas of life, some involve foreign diplomacy, others congressional votes. From big tech announcements to an international award.
On 26 December, the Ukrainian president, Volodymyr Zelenskyy, posted on X that he and Trump would meet before the new year. A cluster of five new Polymarket users cashed in a collective $154,000 from the statement. The bets on “Will Trump meet with Zelenskyy by December 31?” were lodged within a few hours of one another, all one day before Zelenskyy’s statement.
Blockchain data suggests the accounts may be jointly owned. Three of the wallets funding these accounts have transactions between them, and four link to the same Binance wallet. The wallets frequently interact with a Ukrainian-founded cryptocurrency exchange and one features the Ukrainian flag with the bio “#standwithUkraine”.
Earlier this month, Ukrainian regulators banned Polymarket for conducting an unlicensed gambling business.
In October, Venezuelan opposition leader María Corina Machado Parisca won the Nobel peace prize. Shortly after, major news outlets reported about a spike in Polymarket activity before the announcement, with one user having profited more than $65,000, prompting the Nobel Institute to launch an investigation.
But the scale was far bigger than previously reported: the Guardian uncovered eight jointly owned, new accounts which collectively profited more than $161,000, with wagers placed 12 hours before the winner was revealed. Among them are politically charged usernames like “fmaduro”, “madurowilllose”, “trumpdeservesit” and “striketheboats”.
One new account made $17,000 betting on the date of the US government shutdown, the bulk of which was made from a bet made the night before the House passed the government funding bill.
Another made $25,000 predicting which shows would be Netflix’s first and second most-streamed shows that week – in the US and globally – with correct bets placed in 10 Netflix markets for two weeks in a row.
Three have bet exclusively on announcements from Strategy, a publicly traded bitcoin treasury company, profiting more than $143,000 since July.
Twelve emerged in October to bet on the launch of OpenAI’s new browser, profiting $50,000, with all but one now lying inactive.
Could it be that all of these accounts were just normal users who randomly placed life-changing amounts of money on niche markets, took the profits, and never gambled again? It’s conceivable – but seems almost impossible. The likelihood is that these bets were made by people with prior knowledge of the market’s outcome.
The problem is proving it. Because what constitutes insider knowledge is ambiguous.
Take this macabre example: a new Polymarket account made $142,000 betting that Thailand would strike Cambodia just hours before the market resolved. Blockchain data links the wallet to a person based in Thailand. But reports suggest the bets occurred while the strikes had already begun. Seeing missiles, getting evacuated, or hearing local reports first does not make them an insider – they simply had the advantage of circumstance.
Whether insider trading should be illegal, even in more traditional markets, was for a long time a matter of debate. Up until the 1930s, this kind of activity was allowed on the stock market; corporate insiders could freely trade on non-public information. But after the 1929 Wall Street crash, in order to keep ordinary Americans trading, Congress needed to convince them the stock market was not a rigged casino. So, in 1934, it outlawed “any manipulative or deceptive device” in connection with trading securities.
Regulators and courts have since interpreted that law to include trading on the basis of “material, non-public information”. There’s been thousands of cases over the decades, including a lawyer sentenced to 12 years in prison for trading on information stolen from pre-eminent law firms.
Kalshi, Polymarket’s main competitor, has rules prohibiting insider trading. A spokesperson said they freeze accounts that show suspicious behavior while they investigate trades by analyzing that user’s social connections and have already “disciplined a number of users”. Kalshi says that it is a financial crime for users that have “material, non-public information”, about a market to place a bet on that market.
Polymarket is not centrally regulated and does not ban insider trading. Instead, it falls to the users to comply with the law wherever they are based, a spokesperson said.
However, after the CFTC allowed Polymarket to operate in the US in November, Polymarket rolled out a regulated, waitlist-only version of its app for US users. This app does require a government-issued ID and prohibits insider trading. It only has a limited selection of markets.
But US users only need a VPN to access the unregulated version of the site and it’s presumed millions of Americans have placed hundreds of millions worth of bets in this way. This violates Polymarket’s terms of use, but because Polymarket does not collect customers’ personal information, it can only verify their identity with an IP address, which can easily be altered with a VPN.
“Of course there’s people who are working on it that know when it’s going to come. What’s cool about Polymarket is that it creates this financial incentive for people to go and divulge the information to the market and the market to change,” CEO Coplan told 60 Minutes.
“Insider trading is a core feature of the system. It’s what makes them the fastest and most accurate news source in the world,” said Andrew 10 Gwei, who has been exposing insider activity within crypto startups since 2020 and has more recently turned his focus to Polymarket. He estimates to have found dozens of these accounts over the past month or two, totalling millions of dollars in profits.
X is inundated with self-proclaimed insider-hunters like Andrew who track suspicious activity and use this information to inform their own trades. Andrew lets people know when he thinks he has spotted an insider, so that laypeople can quickly copy their bets.
“People are keen to learn how to track insider accounts, or at least get information about them, primarily to find financial opportunities. I doubt anyone following me considers insider trading on Polymarket wrong or harmful,” said Andrew.
Rather than insider traders, Andrew likes to describe the people betting on Polymarket with near-certainty as “privileged money flow” – people who possess unique expertise and positioning that allows them to predict the correct outcome of specific events before everyone else.
“I see nothing wrong with, say, a meteorologist with 20 years of experience and access to modern equipment earning an extra salary on Polymarket. Because of their position, they can predict weather changes with 99% more accuracy than the rest of the market,” he said.
But this mechanism also gives a financial incentive to divulge classified material.
“If you know we’re going to bomb Iran in the next week and you start placing bets, then the prediction market tells the Iranians they’re about to get bombed,” said Richard Painter, who was the chief White House ethics lawyer under George W Bush.
So, in theory, Maduro could have escaped if he had simply checked Polymarket.
“That could be a very, very dangerous thing, because all the countries need to do is look at the prediction markets and try and figure out what the heck we’re going to do,” he said.
A Polymarket spokesperson said that the CFTC-regulated version of its app, which requires government-issued ID and prohibits insider trader in compliance with the law, was currently being rolled out for US users. However, the main site – where the markets the Guardian identified were created – only requires a crypto wallet. While it does not ban insider trading, local laws still apply to the markets, the spokesperson said. Accessing this site via VPN to circumvent geoblocking violated Polymarket’s terms of use, they added.
The military map that changed to win a bet
The claim that insider trading makes markets more honest has a long history in libertarian thought. In 1966, the prominent US legal scholar and economist Henry Manne argued insider trading would make the market more efficient, and letting insiders buy stock before good news goes public could function like a performance bonus.
But the law went in the other direction. From the 1960s, the Securities and Exchange Commission (SEC) aggressively expanded insider trading enforcement, warning that it defrauds investors, reduces trust in markets and distorts corporate behavior.
The same harms apply to prediction markets, experts warn.
Let’s say you bet $50 that Sinners will win best picture at the 2026 Oscars. Hamnet takes it. Your losses are used to pay the winning wagers.
Except, let’s say one of those winners works at PwC, the consultancy that organizes the voting. Suddenly, it no longer feels like a loss, but a theft.
It’s clear that fortune is not evenly distributed on Polymarket. Fewer than 0.04% of Polymarket accounts have profited a collective $3.7bn, capturing more than 70% of total realized profits, data from the blockchain analyst DeFi Oasis shows.
On Polymarket, the comment sections beneath markets with sudden, unexpected outcomes are littered with disgruntled users suspecting foul play.
If enough traders start to feel the markets are rigged, an “unfairness epidemic” can take hold, said Yadav, meaning less trading volume and less accuracy.
But the risks run deeper than investor confidence. In SEC v Texas Gulf Sulphur Co (1968), the court warned that insider trading laws are needed to prevent corporate management from “manipulating the timing and flow of information for personal advantage”.
Prediction markets also risk distorting real-world decision making – but with significantly higher stakes.
On 7 January, Kalshi showed a 98% probability that White House press secretary Karoline Leavitt’s briefing would exceed 65 minutes. But then, with 30 seconds to go, Leavitt concluded abruptly. Winning traders saw returns of nearly 5,000%.
While this may have been coincidence, Painter warned the stakes could get higher, like someone in the Department of Justice recommending a pardon because their friend has a bet on it.
“The real damage isn’t the people who were dumb enough to place a bet. The real damage to the public is a government official who makes a decision they otherwise wouldn’t have made,” said Painter. “That could have a cataclysmic effect on our government and the ethics of our government.”
We have already seen an example of someone in a position of authority reshaping the world to win a bet.
One market speculated on what date Russia would capture the Ukrainian city of Myrnohrad. It was resolved on 15 November when the live map generated by the Institute for the Study of War (ISW), which Polymarket uses to adjudicate territories, updated to show a Russian advance. Then, moments after the market closed, the map reversed and the advance disappeared. The Guardian identified three new accounts that made large bets either the night before the map changed, or in the minutes after, profiting $9,300 in total – a return of about 3,500%.
The ISW’s “control of terrain” map is used by journalists, governments and analysts to track the frontlines of the Russia-Ukraine war. Manipulating it spreads disinformation about a Russian attack and could interfere with the conflict itself.
The ISW shared a statement on its website acknowledging that an “unauthorized and unapproved” edit of its map had taken place, but did not mention betting markets. A spokesperson said the ISW “strenuously objects” to Polymarket using its map, and it disapproves of all betting on the Ukraine war.
Just days after the Maduro trade, Representative Ritchie Torres of New York introduced the Public Integrity in Financial Prediction Markets Act banning members of Congress, their aides or administration officials from making trades on prediction sites based on “material, non-public information”.
“The CFTC’s authority over prediction markets is poorly defined. It’s a gray area. There’s a need for legislative clarity from Congress,” Torres said. With 39 Democratic co-sponsors already, he said he was optimistic about finding a Republican co-lead.
Torres considers the bill a “starting point” for a broader regulatory framework for prediction markets, including a ban on insider trading. Outlawing government officials from trading on insider information is most urgent, he said, as bets create a “perverse incentive” to influence political decisions.
The bill, however, falls short of a total ban of government officials using prediction markets. That means even if it passes before November’s midterm elections, betting on the results may be fair game.
“But you can intentionally lose an election,” Painter points out, who advocates prohibiting officials from these sites altogether.
As Torres put it himself: “When you’re a government insider, the term prediction market is a misnomer, because if you’re the one making the decision, or if you’re part of the decision-making process, you’re not predicting anything. You’re governing for profit.”
