Any Other Uncertain Event

Tarek Mansour spent, by his own account, most of the first four years of Kalshi's existence not building a product but arguing with a regulator. He went to the Commodity Futures Trading Commission before the company had a single live market and asked, essentially, how do we do this legally. The regulator's answer rested on two pillars: market integrity and customer protection. The rest of the answer took a lawsuit against the regulator itself, and court rulings that arrived (he notes with some satisfaction) just in time for the 2024 presidential election. Four years of unglamorous, expensive regulatory patience bought Kalshi something specific: a jurisdiction. Not gambling law.

Commodities law.

A yes/no contract on whether inflation clears a threshold is, under that jurisdiction, a derivative (the same species of instrument as a wheat future) and derivatives are regulated by people who have never once asked whether the trader predicting the harvest is lucky or merely well-informed.

That distinction is the whole company. It is also proving to be a distinctly American one.

The list, so far

France blocked Polymarket in 2024; Belgium, Germany and Poland followed. On 20 January 2026 two regulators moved on the same day. Portugal's SRIJ ordered Polymarket to stop operating, and its reasoning is the one worth keeping: the site is not authorised to offer betting, and under national law betting on political events or happenings, whether national or international, is not permitted. That is a European regulator looking directly at a Kalshi-shaped contract and reaching for gambling law without a second thought. Hungary's supervisory authority blocked the domain and its subdomains the same day, for "forbidden organization of gambling activities". Polymarket's own list of blocked jurisdictions adds the United Kingdom. Then, on 26 May 2026, Spain's Dirección General de Ordenación del Juego ordered every Spanish ISP to block both platforms at the DNS and IP level and opened formal proceedings.

Eight jurisdictions, reaching the same result through separate national processes, months apart, with no coordination between them. Trade coverage of the Spanish block cites a longer tail still - Switzerland, Romania, the Netherlands - though without individual sourcing for any of them, which is its own small, honest footnote: once enough jurisdictions start quietly doing the same thing, even the people counting them stop being sure exactly how many there are.

Not one of them held a hearing on whether skill predominates over chance. Not one needed to.

And the Spanish notice says why, in words the drafters of this piece could not have improved on

The DGOJ's stated ground is that the platforms were offering betting products tied to uncertain future events without the licence Spanish law requires. It is worth reading that slowly, which is the one way regulatory notices are never expected to be read. Not products whose outcome turns on luck rather than research. Not games in which skill fails to predominate. Products tied to uncertain future events - a phrase which, taken literally, would also describe a wheat future, a weather derivative, a pension fund and most of agriculture. Somewhere in Madrid, a civil servant with a queue of other files and no particular interest in American commodities law wrote, in one subordinate clause, the problem Kalshi had spent four years and a federal lawsuit arguing its way around. Nobody at the DGOJ was trying to settle whether prediction is a skill. They simply described the thing in front of them, and the description turned out to have no room for the question.

Two honest qualifications before that gets pushed too far. The DGOJ did not rest on the category alone: it also cited the absence of identity verification, of protections for minors, and of systems for self-excluded gamblers (substantive consumer-protection grounds that have nothing to do with taxonomy and would be worth answering on their own terms). And the Spanish block is precautionary, in place while a proceeding expected to run three to four months reaches a conclusion; Hungary's is likewise temporary, pending that authority's own review.

Which is, if anything, the sharper observation. In at least two of these jurisdictions the substantive question is formally still open. The platforms are already off. Nobody is hurrying. A provisional measure that works perfectly well is not under much pressure to become a decision, and a question that everyone has managed without answering tends to stay unanswered.

Europe's second door is locked from the other side

Kalshi's entire regulatory triumph rests on convincing one federal commodities regulator that it belongs to commodities law instead of gambling law. Europe does have a door of sorts: MiFID II, which is why a European wheat future answers to a financial regulator rather than a gambling one, and whose list of financial instruments reaches, on paper, derivatives relating to inflation rates and other official economic statistics. But that door opens only for someone who arrives holding a national financial regulator's authorisation, and none of the blocks above was issued against a platform that did. There is no EU-level commodities regulator with a plausible claim on event contracts to hold it open; gambling regulation is not harmonised at EU level at all; it sits with each member state, a competence the Court of Justice has repeatedly confirmed states may exercise as they see fit - and every regulator that has looked at Kalshi or Polymarket so far has looked through the only door available. Not because they examined the skill question and decided against Kalshi. Because the question of skill was never on the form.

Croatia has not blocked either platform. Its statute already explains why it would not need to think hard about it if it did

Article 2 of the Zakon o igrama na sreću defines a game of chance as one in which, in exchange for a stake, a participant may win money, goods, services or rights, where the win or loss depends predominantly on chance or on any other uncertain event.

The predominance test is there - pretežito - and it is worth being precise about, because a careless reading of this provision invites an easy rebuttal. Croatian law does not dispense with predominance. It changes what predominance is measured against. The American test asks whether skill predominates over chance: establish that your traders are researchers rather than gamblers and you are outside the category, which is the load-bearing finding Kalshi spent four years and a federal lawsuit securing. The Croatian test asks whether the outcome depends predominantly on chance or on any other uncertain event - and the second limb swallows the argument whole. A contract on next month's Fed funds rate depends predominantly on an uncertain event no matter how much research the trader brings, because the research does not make the Federal Open Market Committee's decision any less uncertain at the moment the stake is placed. There is no showing a well-informed trader could make that takes the contract outside the definition, because the definition never asked how well informed anyone was. (Well, strictly speaking, there is one exception: the trader who already knows how the Committee will vote. For that trader the event is not uncertain at all, and the contract is not a game of chance. It is simply a different offence.)

This is not a translation gap. It is a different design

Croatia's drafters were almost certainly not thinking about prediction markets when they wrote "any other uncertain event" - the phrase reads like it was aimed at prop bets and novelty wagers, the century-old genre of betting on which raindrop reaches the bottom of the window first, not at a Federal Reserve contract. But statutes do not stay aimed at what their drafters were looking at. They cover what the words reach, and "any other uncertain event" reaches considerably further than its drafters needed it to. A test built around why nobody knows the answer would have left room for information markets to argue their way out, the way Kalshi did in Washington. A test built around the plain fact that nobody knows the answer leaves no such room. The gap is not that Europe forgot to write an exception for skill-based forecasting. It is that the machinery never asked the question that would have required one.

Nor is the American door simply open

It is worth resisting the version of this story in which the United States permits and Europe forbids. Under 17 CFR 40.11 the CFTC restricts whole categories of event contract outright: terrorism, assassination, war, gaming, anything unlawful under state or federal law, anything contrary to the public interest. One of those categories is literally gaming. The American system is not unbounded; it is bounded differently, and (this is the part that matters) Kalshi was in the room while the boundary was being drawn around it, for four years, at considerable expense. That is the asymmetry. Not permission versus prohibition, but whether the operator gets to participate in deciding which side of the line it falls on.

The part that makes this a small question rather than a large one

None of this is really about protecting Croatian citizens from a novel form of gambling harm, and the statute does not pretend otherwise. Article 9 hands the lottery category to Hrvatska Lutrija outright - a state-owned monopoly, untouched by any of this reasoning, selling structurally identical uncertain-event products under a different Article. Everything else (casinos, betting shops, the categories a private operator might conceivably enter) requires not a licence application processed against fixed criteria, the way the CFTC clears a new contract by the afternoon unless it intervenes, but a discretionary decision of the Government of the Republic of Croatia. One is a form. The other is politics. A well-capitalised, well-lawyered platform can, in principle, survive four years of American regulatory argument and come out the other side a business. Surviving four years of "the Government has not yet decided" is a different kind of company entirely - one that mostly does not get built, which is a way of saying the incumbent's position was never actually threatened.

The penalty schedule is almost an afterthought next to that, though it is not small: €6,630 to €66,360 for a company that organises unlicensed games of chance, €660 to €6,630 for the individual who signed off on it. Enough to matter to a startup. Not the actual barrier. The actual barrier is upstream of the fine - it is the absence of a form to fill out.

The Clause has read a great many definitions of uncertainty...

...and it has never once cared whether the uncertainty in question was worth insuring against, worth hedging, worth studying, or worth nothing at all beyond the thrill of not knowing. "Any other uncertain event" was, in the fullness of time, always going to reach further than the raindrop on the window. Nobody has yet been paid to go back and ask whether it was supposed to.

Kalshi's American story is usually told as a company that fought a regulator and won. The more useful reading, from this side of the Atlantic, is narrower: it fought one specific regulator, in one specific jurisdiction, that happened to have a second door in its wall and a willingness to argue about which room the company belonged in. Most jurisdictions have neither. The interesting question this site keeps returning to is not whether prediction markets are good policy - reasonable people disagree, and this piece has not tried to settle it - but how much of what looks like Europe rejecting a technology is actually just Europe never having built the door it would have needed to let it in.