The prediction market feud looks like two founders who hate each other. Underneath it is a business model fight over whether a license is a cost or a moat.
Kalshi and Polymarket built the same product on opposite legal foundations. Kalshi spent years and millions getting licensed in the United States, then used that position to lobby regulators and prosecutors against its rival. Polymarket launched offshore and grew faster. Kalshi is now worth $22 billion, about $7 billion more.
Somewhere in a Manhattan conference room in 2024, a group of lawyers sat down with federal prosecutors and explained, helpfully, how a competitor’s website worked. Not a whistleblower. Not a regulator. A rival company’s outside counsel, walking the government through the mechanics of a product it wanted shut down.
That is the part of this story worth your attention, and it is not really about two men who cannot stand each other.
What Happened
The New York Times reported this weekend on the feud between Shayne Coplan of Polymarket and Tarek Mansour of Kalshi, built on interviews with more than 40 people in their circles. The paper’s news line is that Kalshi’s lawyers met with federal prosecutors in Manhattan in the months before the FBI raided Coplan’s apartment in November 2024, flagging that Americans could reach Polymarket despite the ban.
Kalshi’s spokeswoman says the company learned about the raid from media reports and that it routinely discusses compliance with the government. Polymarket’s response to a request for comment on Mansour was a single line of Emerson: the louder a man talks of his honor, the faster you count your spoons.
The two companies have filed competing trademark applications, chased the same partnerships, poached each other’s staff, and both cultivated ties to Donald Trump Jr. Kalshi built an industry advocacy group to fight state gambling regulators and pointedly did not invite Polymarket. When Kalshi lost a state case, Polymarket opened a market on how fast Kalshi would have to shut down its sports book in Massachusetts.
The Backstory
Mansour, a former Citadel trader, started Kalshi in 2018 with MIT classmate Luana Lopes Lara and went to the Commodity Futures Trading Commission for permission before taking a single bet. Coplan launched Polymarket in 2020 out of a Lower Manhattan apartment and did not.
The gap compounded from there. In January 2022 the CFTC fined Polymarket $1.4 million for running an unregistered platform, and the company agreed to stop serving U.S. customers. Kalshi had been raising concerns about Polymarket with the agency since its own licensing discussions, according to the Times.
Then the plan backfired. The ban was supposed to clear the U.S. field for Kalshi, but the CFTC also blocked bets on elections, and Kalshi spent 2023 and 2024 suing the agency instead of selling. Polymarket, banned but unbothered, became the story of the 2024 election as Americans reached it through VPNs. Kalshi won its case shortly before the vote. Coplan got the FBI at his door the week after it.
The Plan
Both companies have since converged on the same destination from different directions.
Polymarket bought its way onshore. After the Southern District dropped its investigation and the enforcement climate shifted, Coplan acquired QCX, a licensed exchange, for $112 million, and launched a smaller compliant U.S. app while the main offshore platform kept running. Intercontinental Exchange, the owner of the New York Stock Exchange, committed up to $2 billion, reportedly after Mansour and a Kalshi investor made a late push to talk ICE out of it.
Kalshi went the other way, converting its regulatory standing into consumer distribution: a Madison Square Garden partnership complete with a branded concourse, sports contracts, retail brokerage integrations. A former CFTC commissioner sits on its board and in May posted publicly about Polymarket’s user screening, tagging the U.S. Attorney’s office for the Southern District of New York.
Both are now priced for dominance. Kalshi is reportedly in talks for a round at roughly $40 billion, which would nearly double its last valuation. Polymarket has been raising at around $15 billion.
The Business Model Angle
Here is the thing most coverage of this feud will miss. The interesting question is not who is more likeable. It is what a license actually buys you.
For most of the last decade, the standard startup playbook treated regulation as a tax to be minimized and, ideally, deferred. Launch, scale, apologize, lobby. Uber ran it. Airbnb ran it. Polymarket ran a purer version of it than either, by simply operating from Panama and letting the product find its users.
Kalshi ran the opposite playbook, and the crucial part is that it did not treat compliance as a defensive cost. It treated it as an offensive asset with three distinct returns.
The first is exclusion. A licensed incumbent can credibly ask regulators to act against an unlicensed rival, and that request lands very differently coming from a company that has already done the paperwork. Kalshi’s lawyers were not lying to prosecutors. They were describing a real gap, which is exactly what made the tactic effective. Compliance bought Kalshi standing to be believed.
The second is distribution. Licensed exchanges can sign partners that offshore platforms cannot touch. An arena, a broadcaster, a brokerage, a bank: all of them have their own compliance departments, and those departments do not clear counterparties operating from Panama. That is why Kalshi got the Garden and Polymarket did not, and it matters far more than the $7 billion valuation gap. The valuation is an opinion. The partnership pipeline is a structural advantage that compounds.
The third is identity data. Kalshi requires customers to hand over personal details, sometimes including employer, before they trade. That is friction, and friction costs signups. It also produces a proprietary dataset and a defensible answer when a regulator asks how the platform prevents insider trading. Polymarket’s answer is different and genuinely clever, since every trade is public and can be audited by anyone, but it is an answer about detection rather than prevention. In April an Army Special Forces soldier was charged with betting on the capture of Venezuela’s president using confidential information, clearing more than $400,000 on Polymarket’s offshore platform through a VPN.
For founders in any regulated adjacency, that is the transferable lesson. The permissionless playbook optimizes for speed of user acquisition. The permissioned playbook optimizes for who is legally allowed to do business with you. In a market where the biggest revenue unlocks come from institutional partners rather than individual signups, the second one wins, and it wins later.
The Risk
The tidy version of this story is that compliance beat arbitrage. Be careful with it.
Polymarket is not losing. It raised from the owner of the New York Stock Exchange, escaped federal prosecution, bought a license, and still runs the larger global user base. If ICE’s real thesis is data distribution rather than betting volume, then Polymarket’s public order book is the asset, and Kalshi’s KYC vault is worth less than it looks.
Kalshi’s moat also has a leak. Its advantage is regulatory, and regulation moves. Both companies are now fighting state attorneys general who argue their sports contracts are gambling by another name, and Kalshi has already lost ground in at least one state. A federal license does not settle a state gambling question, which means the very thing Kalshi paid for is the thing being litigated.
Then there is the aggression itself. Both companies have crossed lines, as a former Kalshi employee told the Times, and that reads badly to the institutions both now need. A Kalshi employee once directed a former NFL player to post that Coplan was guilty. Polymarket’s influencers amplified news that a Kalshi partnership had collapsed. That behavior is affordable when you are a crypto curiosity. It is expensive when you are asking Morgan Stanley and Madison Square Garden to underwrite your legitimacy.
And both are priced as if the category is settled. Combined trades passed $150 billion in the first half of 2026, up roughly 1,200 percent, which is the kind of growth that invites entrants. Meta is reportedly building a prediction markets app. That is a distribution player with three billion users walking into a market where distribution is the moat.
Quick Questions
Which is bigger, Kalshi or Polymarket? Kalshi, on both measures that currently matter. It was valued at $22 billion at its last completed round, roughly $7 billion above Polymarket, and processed more than $33 billion in trades in the most recent full month, about twice Polymarket’s total according to The Block.
Is Polymarket legal in the United States? Partially. Its main platform still operates offshore and remains barred from serving U.S. customers, but the company acquired QCX, a CFTC-licensed exchange, for $112 million and runs a smaller compliant American app on that license.
Why does Kalshi have a regulatory advantage? It sought CFTC approval before launching and later beat the agency in court over election contracts, which established the legal pathway for licensed prediction markets in the United States. That standing lets it sign partners whose compliance teams will not clear an offshore counterparty.
What is the actual business model difference? Kalshi is a permissioned exchange: identity verification up front, restricted market categories, institutional distribution. Polymarket is a permissionless one: anonymous signup, a public on-chain order book, near-unlimited market topics. One optimizes for who will do business with it. The other optimizes for how fast it can grow.
Are prediction markets profitable? Both take a cut of transaction activity rather than betting against customers, so revenue scales with volume rather than with user losses. At current valuations against current volumes, both are priced well ahead of demonstrated earnings.
The Business Model Analyst Take
Strip out the courtside photos and the Emerson quotes and this is a clean natural experiment. Two teams, same product, same market, same eighteen months of explosive demand. One paid the regulatory toll up front. One did not.
The scoreboard right now favors the one that paid, but the reason is worth getting right, because the wrong lesson here is “follow the rules and you will win.” Kalshi is not ahead because regulators rewarded virtue. It is ahead because a license is a key to a room full of counterparties, and those counterparties control the distribution that turns a niche product into a habit. The compliance spend was never really a legal expense. It was customer acquisition, billed to the wrong department.
If you are building anywhere near a regulated market, the question is not whether to comply. It is whether compliance in your category unlocks partners you cannot otherwise reach. If it does, the toll is a moat and you should be paying it early and loudly. If it does not, you are just buying permission, and permission is not a strategy.
Reporting on the Coplan and Mansour rivalry, the meetings with federal prosecutors, and the valuation and volume figures is from The New York Times, “Do These C.E.O.s Loathe Each Other? Bet on It,” by David Yaffe-Bellany and William K. Rashbaum, July 19, 2026.
