Banned in Thirteen Countries

On January 18, 2026, Portugal held a presidential election. Polymarket recorded over EUR 103 million in wagers on the outcome. Within two hours of official results, EUR 4 million traded in a single burst. Two days later, the Portuguese Gaming Regulatory Authority issued a 48-hour operational halt directive and blocked the platform.

The volume was overwhelmingly placed from outside Portugal, on a platform headquartered in New York, settled in USDC on Polygon, priced in contracts denominated in US dollars. Portugal's regulators were explicit: the website was not authorized to offer betting in Portugal, and under national law, betting on political events is not permitted.

What makes Portugal instructive is the sequence. The prediction market did not just measure the election. The volume it generated on the election provoked its own prohibition. The market price became an input to the regulatory environment that governs the market. This is reflexivity applied to regulation itself.

The ban chronology

Polymarket is now geo-blocked in 33 countries and formally banned or restricted in at least 13. The timeline:

November 2024. France moves first. The Autorite Nationale des Jeux classifies Polymarket as unauthorized gambling, with fines of up to EUR 200,000 and potential prison sentences for operators. Switzerland's Gespa orders ISP-level blocking. Polymarket implements voluntary geo-blocks in both countries, retaining view-only access for French users: charts, odds, and historical data visible, but trading disabled.

January 2025. Poland adds Polymarket to its registry of prohibited gambling websites. Singapore's Gambling Regulatory Authority blocks access.

February 2025. Belgium follows.

August 2025. Australia's ACMA orders ISP-level blocking after an investigation finds the platform was paying TikTok influencers to promote betting during the May federal election. Polymarket joins ACMA's blocked list alongside offshore poker rooms and unlicensed slot sites.

Late 2025. Romania's National Gambling Office blacklists Polymarket after it hosts wagers on the Romanian presidential election. Hungary's regulator blocks all Polymarket domains and subdomains. Users with Hungarian IP addresses see a regulatory warning in place of the site.

January 2026. Ukraine blocks Polymarket under an illegal gambling crackdown. Portugal issues its halt directive after the presidential election volume spike.

The full restriction list as of February 2026: France, Belgium, Singapore, Thailand, Switzerland, Poland, Australia, Romania, Hungary, Portugal, Ukraine, Italy, and the United Kingdom. Thirty-three countries geo-blocked. The United States was blocked until late 2025, when Polymarket began a waitlisted reentry.

Approximately 180 countries remain technically accessible. But in many of those jurisdictions, Polymarket operates without local authorization. The platform exists in a gray zone between accessible and legal.

Four regulatory frameworks

The global response does not follow a single logic.

The United States treats prediction markets as financial instruments. The CFTC regulates event contracts as commodity derivatives. Kalshi and Polymarket both hold Designated Contract Market status. New CFTC Chairman Michael Selig, in his first public remarks on January 29, 2026, withdrew the proposed ban on sports and political event contracts: “It is time for clear rules and a clear understanding that the CFTC supports lawful innovation in these markets.” The federal posture is increasingly permissive. The state posture is not. Nevada filed a civil enforcement action against Polymarket. New York's attorney general called prediction market products bets “masquerading” as event contracts. The federal-state conflict is heading toward the courts.

Europe treats prediction markets as gambling, with no unified framework. Individual member states classify and ban independently. France, Belgium, Poland, Italy, Romania, Hungary, and Portugal have all blocked or restricted Polymarket under national gambling laws. Germany and Spain still allow access. The Markets in Crypto-Assets regulation addresses crypto-asset services broadly but does not mention prediction markets specifically. Austria, France, and Italy have called for a stronger European framework. The MiCA grandfathering period ends July 2026. After that date, any crypto-native platform operating without a local license faces stricter compliance burdens.

The United Kingdom treats prediction markets as a licensed gambling product. The Financial Conduct Authority deferred to the Gambling Commission. Matchbook launched the UK's first licensed prediction market in January 2026, built on its existing UKGC exchange licence. The UK model is clear: prediction markets require a gambling license, and crypto-native platforms are effectively excluded by FSMA rules, AML requirements, and advertising restrictions.

Asia defaults to prohibition. Singapore, Thailand, Taiwan, and China have blocked platforms. South Korea's gambling law is prohibition-first with narrow state-run exceptions. Taiwan was the first jurisdiction globally to call for blocking Polymarket, triggered by betting on its 2024 presidential election. There is no regulatory pathway for prediction markets as financial instruments in most Asian jurisdictions.

Four frameworks. No international coordination mechanism. No body — not IOSCO, not the FSB, not FATF — has attempted to harmonize cross-border prediction market regulation.

The arbitrage geography

While Polymarket accumulates bans, Kalshi has been accumulating countries.

Kalshi launched globally in October 2025, expanding to over 140 countries alongside a $300 million Series D at a $5 billion valuation. The structural choice was significant: Kalshi operates a single unified liquidity pool. International users access the same order books, the same prices, the same liquidity depth as American users.

By early January 2026, Kalshi handled roughly two-thirds of total global prediction market trades. On a single day in January, the platform processed approximately $466 million. During the week ending January 11, it recorded over $2 billion in weekly notional volume for the first time.

Kalshi restricts over 50 countries, including the UK, Canada, France, and Singapore. But its 140-plus accessible countries are meaningfully more than Polymarket's effective reach. And the unified liquidity pool means price discovery in accessible jurisdictions sets the global benchmark.

This is where the geography becomes reflexive.

When Portugal bans Polymarket, it does not eliminate prediction market data about Portuguese elections. It eliminates Portuguese participation in setting that data. The price signal existed before the ban. It exists after the ban. Portuguese voters, media, and analysts consumed it regardless. Portuguese regulators could ban the transaction but not the information.

The pattern holds across every banned jurisdiction. French media cited Polymarket's odds on the 2024 US election extensively — after France banned the platform. Singapore's financial press reports Kalshi prices. Australian journalists reference prediction market probabilities on their own elections, sourced from platforms their government has blocked.

Every ban concentrates price-setting power in the jurisdictions that remain open. The US and its CFTC-regulated platforms are becoming the de facto global price-setters for event probability. Polymarket settles trades on a public blockchain accessible via VPN from anywhere. Kalshi operates a single liquidity pool across 140-plus countries. The banning jurisdictions are not opting out of prediction markets. They are opting out of participation in prediction markets.

You can ban the market. You cannot ban the price signal.

The institutional acceleration

The data distribution infrastructure does not respect national boundaries.

Intercontinental Exchange has invested in Polymarket with a deal to become the exclusive global distributor of Polymarket's real-time event-driven data. Dow Jones integrates Polymarket probabilities into the Wall Street Journal, Barron's, and MarketWatch. Google Search displays live odds from both Polymarket and Kalshi.

A Portuguese journalist searching Google for election probabilities will find Polymarket data in the results — served by Google, sourced from a platform Portugal has banned. A French analyst reading the Wall Street Journal will encounter Polymarket odds on the French fiscal outlook — distributed by Dow Jones, produced on a platform France prohibits its citizens from using.

The information layer is borderless. The transaction layer is not. This asymmetry is the structural fact of global prediction market regulation in 2026. And it creates a feedback loop that the banning jurisdictions have not reckoned with: the more countries ban prediction market platforms, the fewer jurisdictions participate in price-setting, the more concentrated price-setting power becomes in the US and other permissive jurisdictions, the more authoritative those prices appear, the more widely they get cited, the more they shape outcomes in the very countries that imposed the bans.

The reflexive regulatory loop

The Portugal case illustrates the general pattern.

Prediction markets attract volume on politically sensitive events. Volume creates visibility. Visibility triggers regulatory scrutiny. Scrutiny produces bans. Bans push volume to platforms in permissive jurisdictions. Concentrated volume deepens liquidity and improves price discovery. Better prices get cited more widely. Wider citation increases visibility in the jurisdictions that imposed bans.

This is the observable sequence in every country that has banned Polymarket. Romania blacklisted the platform after wagers on its presidential election. Taiwan called for a block after betting on its presidential race. Australia's investigation was triggered by election-adjacent promotion. The political sensitivity of the event drives the volume. The volume drives the ban. The ban does not drive the volume to zero. It drives the volume somewhere else.

The loop has a second-order consequence that regulators have not yet confronted. When a country bans its citizens from trading on prediction markets, it does not reduce the influence of prediction market prices on that country's politics, elections, or public discourse. It increases it — because the prices are now set entirely by outsiders, with no local participation to correct for local knowledge, local context, or local sentiment. The ban does not eliminate the reflexive loop. It makes it worse.

Three implications

The regulatory map is a liquidity map. Every ban reshuffles where volume concentrates. As Europe fragments — Germany and Spain open, France and Poland closed, MiCA grandfathering ending July 2026 — the jurisdictions that remain accessible will absorb disproportionate volume. Kalshi's two-thirds global market share is not just a product of its platform. It is a product of the regulatory geography.

Prediction market data is becoming a public good produced by private markets in permissive jurisdictions. When Google Search displays Polymarket odds alongside Wikipedia summaries and news articles, the probability estimate functions as infrastructure. It informs decisions in every country, including the ones that banned the platform. The countries that allow prediction markets are not just hosting transactions. They are hosting the production of a global information resource.

The industry has not priced its own most important risk. The reflexive loop between prediction markets and their own regulation is more consequential than sports volume growth, TradFi integration, or blockchain scalability. The industry's trajectory depends on whether federal preemption holds in the US, whether MiCA's July 2026 deadline triggers a wave of European restrictions, and whether any international body attempts coordination. These are not background conditions. They are the foreground. Prediction markets exist to price future events. They have not yet built adequate markets on the event that matters most to their own survival: their regulatory trajectory.

Thirteen countries have banned Polymarket. Thirty-three have geo-blocked it. The industry did $40 billion in volume in 2025 anyway. The bans are real. The volume is also real. Both facts are true simultaneously, and the tension between them is the defining structural condition of the prediction market industry in 2026.

Dawson Smith writes Reflexivity, a newsletter on prediction markets as reflexive systems. Research and analysis, not trading advice.

dawson@monetizeopinion.com