A new analysis published yesterday estimates that between $11 billion and $34 billion of offshore prediction market volume came from US users. The upper end of that range has already done the rounds in headlines - "US users secretly wagered $34 billion" - but the number itself is less important than what it exposes about a system nobody has agreed on how to govern.
The research comes from prediction market analyst Harry Crane, who estimated US user activity across all tracked offshore venues. Total offshore volume in the period he examined was roughly $159 billion, meaning American users account for 7% to 21% of it depending on how you model IP addresses and geographic proxies. That's a range, not a figure. And some of that volume may be inflated - a Columbia University study from November found that nearly 25% of Polymarket's reported trading volume may be fake, driven by wash trading (the practice of repeatedly buying and selling the same position to manufacture activity). I mention the wash trading not to dismiss the whole exercise, but to flag that any volume estimate in this space should be read as directional, not precise.
What the analysis actually reveals is less interesting in its arithmetic and more revealing in its plumbing.
The rails matter more than the restrictions
Polymarket - the largest offshore prediction market by far - has been blocked from the US since a 2022 settlement with the CFTC (the Commodity Futures Trading Commission, which regulates derivatives markets). It says it prohibits VPNs. It says it restricts order placement from US geographic locations. It also runs on a public blockchain, which means every transaction is broadcast on-chain and anyone with the right tools can trace where the money actually moves.
That's the structural paradox. The platform claims geographic compliance, but the settlement layer underneath - Ethereum and its stablecoin rails - doesn't really have a geography. A user in New Jersey can deposit USDC (a dollar-pegged crypto token) and route it through infrastructure that doesn't stop at a border. The geoblock works at the application layer, like a website that declines your browser. It's real, but it's also a thin membrane.
This is worth sitting with for a moment because it's not just a prediction market problem. It's the same question that's been quietly moving to the center of the stablecoin debate: when money moves on rails that aren't designed for jurisdictional boundaries, how does a domestic regulator enforce its perimeter? You can block a website. You can't block a blockchain without blocking a protocol, which is a much heavier lift.
The political pressure is converging
The regulatory environment around these platforms has shifted sharply in the last four months. It's not one agency moving - it's multiple actors arriving at the same target from different directions, and that convergence is the more durable signal than any single enforcement action.
In January 2026, the CFTC withdrew a proposed rule that would have banned political and sports-related event contracts entirely - a surprising reversal that opened a door for regulated platforms like Kalshi to operate. But the door came with a mirror: in April, the SDNY and the CFTC jointly announced enforcement actions against a US Army soldier for insider trading on Polymarket, where he reportedly purchased large positions on war-related outcomes before they became public. That case is significant not because one soldier traded, but because it established the CFTC's working theory that insider trading law applies to prediction markets even when they're offshore.
Then in May, House Oversight Committee Chairman James Comer launched a formal investigation into insider trading on both Polymarket and Kalshi, sending letters to both companies' CEOs. The CFTC's Director of Enforcement has publicly called the idea that insider trading is permissible in prediction markets a myth, and reports from earlier this month suggest the agency is now using AI-driven surveillance to monitor prediction markets globally.
What's happening here is a constituency battle over who gets to define what a prediction market is. Regulators have been moving toward framing them as derivatives markets subject to existing commodity and insider-trading law. If that framing sticks, offshore structuring stops being a shield and becomes an aggravating factor.
Regulated versus unregulated - and why the gap is closing
It's useful to understand the two-track landscape that's now in place.
Kalshi operates under CFTC regulation. US users can deposit through traditional banking, trade event contracts on weather, economics, politics, and - since early 2025 - sports. It's clean on the compliance side, and it's grown to roughly $1 billion in monthly volume. Sports accounts for about 80% of its trading.
Polymarket operates in two modes. The international version runs on-chain with stablecoin deposits, serves global users, and remains banned in the US proper. A CFTC-approved US version launched in February 2026. Sports is about 39% of Polymarket's overall volume - politics and cryptocurrency make up much more of its mix, which is where the insider trading risk lives.
Both are now seeing comparable monthly volumes in the $1 billion range. The gap that existed a year ago - when Polymarket dominated and Kalshi was a niche US alternative - has closed. What hasn't closed is the question of whether the offshore side can keep operating at scale while American regulators treat it as an enforcement problem rather than a market problem.
What changes if the jurisdictional claim holds?
Here's where I think the story gets more structural.
If the CFTC succeeds in establishing that insider trading law applies to offshore prediction markets - and that it has jurisdiction over US participants regardless of the platform's domicile - the implication goes beyond prediction markets. Any on-chain financial product that US citizens can access, but which is hosted offshore and denominated in stablecoins, becomes potentially within the CFTC's reach if the agency is willing to stretch its commodity-futures authority.
That's not a certainty. The CFTC has limits, and platforms like Palantir (which Polymarket has reportedly hired for market surveillance) are making it harder to prove insider patterns without generating a mountain of false positives. But the direction of travel matters.
For an ordinary investor, the implication is narrower than the headline suggests but worth tracking. If offshore prediction markets face sustained enforcement, volume migrates to regulated venues. That helps Kalshi and the regulated track, and it pressures Polymarket's international business model. If enforcement stays sporadic - one soldier, one congressional letter, then silence - the offshore track continues to absorb US demand because the economic incentives (more markets, crypto-native access, broader product range) are still stronger there.
I'm watching whether the CFTC files a second enforcement case against an individual trader on an offshore platform, or whether it moves directly against the platform itself. That distinction would tell us whether this stays a deterrence strategy or becomes a structural one.
The $34 billion number is a headline. The real question is whether US regulators can enforce jurisdiction over a market whose settlement layer doesn't recognize borders - and whether they're willing to try hard enough to find out.