You think a CFTC license is a golden ticket to 50 states. The truth is, one judge in New York just ripped up that map.
On July 26, 2025, the Southern District of New York denied Kalshi's motion to dismiss a lawsuit brought by the New York State Attorney General. The state argued that Kalshi's event contracts on congressional elections violated New York's anti-gambling statutes. Kalshi's defense rested on federal preemption: the Commodity Exchange Act, they claimed, gave the CFTC exclusive authority over all commodity derivatives, including event contracts. The judge disagreed. Her ruling held that nothing in the CEA expressly preempts state gambling laws, and that Congress had not intended to occupy the field. The decision was narrow but devastating. It didn't ban Kalshi—it simply said New York could enforce its own laws against a federally regulated market. The immediate effect: Kalshi's New York operations are now in legal jeopardy. But the ripple effects extend to every prediction market in the United States.
Context Prediction markets allow users to trade contracts based on the outcome of future events—elections, sports, economic indicators. They sit at the intersection of derivatives trading, gambling, and forecasting. Kalshi, founded in 2019, positioned itself as a regulated alternative to offshore platforms like Polymarket. It registered with the CFTC as a designated contract market (DCM). By 2025, Kalshi had processed over $2 billion in volume. The industry narrative was simple: CFTC approval equals legitimacy. The agency had even proposed a new rule in 2024 to clarify which event contracts were permissible. Everyone assumed that once the CFTC signed off, state-level challenges were preempted. That assumption just cracked.
Core Analysis Let's dissect the court's reasoning. The judge applied the two-part test for federal preemption: express preemption and field preemption. The CEA does not contain a clause that explicitly says "state gambling laws do not apply to CFTC-regulated contracts." That's a fatal omission. Courts are loath to infer preemption unless Congress makes it unambiguous. Kalshi argued that the CEA's comprehensive regulatory scheme implied field preemption. The judge rejected this, noting that states have historically regulated gambling, and Congress did not intend to displace that power when it created the CFTC. The ruling effectively creates a dual-sovereignty framework for event contracts: federal approval is necessary but not sufficient.
The practical impact is fragmentation. Prediction markets now face a patchwork of 50 state laws. New York's anti-gambling statutes are strict—it considers any betting on an event of chance or skill as gambling unless specifically exempted. Sports betting is legal in New York only because the state authorized it. Election betting is not exempted. Other states may follow: California, Texas, and Florida have aggressive gambling regulators. The cost of compliance is real. The court explicitly noted that "geographic fencing costs are ordinary compliance burdens," rejecting Kalshi's argument that blocking New York users would be prohibitively expensive. That's a dangerous precedent. It tells every state attorney general that they can sue a federally regulated platform without worrying about the preemption defense.
From my experience auditing financial risk models, I recognize a recurring pattern: the assumption of a single-layer defense. In 2020, I spent weeks stress-testing Compound's interest rate model. I found a rounding error that could lead to infinite yield exploitation under high volatility. The team fixed it, but the lesson stuck: mathematical elegance often masks implementation fragility. Kalshi's business model relied on a single legal assumption—federal preemption. That assumption just collapsed, revealing the fragile architecture underneath. Logic doesn't care about your regulatory strategy; it only cares about the law as it is.
Let's quantify the fragmentation risk. There are 50 states. Assume 60% have gambling laws that could be interpreted to cover event contracts. That's 30 states. Each state requires separate legal analysis, compliance infrastructure, and potentially litigation. The cost per state could range from $500,000 to $2 million annually, depending on legal fees and technical implementation. For a platform like Kalshi with limited revenue, that's a meaningful drag on profitability. More importantly, the uncertainty deters institutional capital. No major fund wants to invest in a platform that could lose access to New York, California, and Texas simultaneously.
CFTC Chair Rostin Behnam has hinted at a forthcoming final rule on event contracts, expected by mid-2026. The proposed rule from 2024 sought to ban certain "political event contracts" outright. But even if the CFTC green-lights some contracts, the state preemption issue remains. The CFTC cannot override state gambling laws unless Congress explicitly grants it the power. That's why the court's ruling is so consequential: it shifts the battlefield from federal rulemaking to state legislatures and courts.
Polymarket, the largest decentralized prediction market, faces even greater exposure. It has no CFTC registration and operates outside the US legally, but it allows US users through VPNs. The New York ruling gives states another tool to go after unregistered platforms. State attorneys general now have a template: sue under state gambling laws, argue no federal preemption, and seek injunctions. Polymarket's users could be targeted individually, but more likely, the state will target any entity facilitating the transaction, including payment processors and hosting providers.
I have seen this movie before. In 2022, I analyzed the Terra Luna collapse through a risk management lens. The trigger was a single large withdrawal from Anchor Protocol, but the root cause was a mispriced risk: the assumption that algorithmic stability would hold under stress. The exploit wasn't in the smart contract; it was in the business model's premise. Kalshi's exploit is the assumption that federal approval equals national access. The trigger is a New York judge. The loss is not $40 billion—it's the entire potential of the US prediction market.
Contrarian Angle Now, let's test the bull case. What did the optimists get right? First, the ruling is not final; Kalshi can appeal. The Second Circuit might reverse, especially if they find that the CEA does occupy the field. The Supreme Court has been skeptical of expansive state gambling laws in other contexts. Second, the ruling is limited to New York. Other states may not sue. The industry can operate in 49 states, which is still a massive market. Third, Congress might step in. Representative Patrick McHenry introduced the "Event Contract Certainty Act" in 2024, which would explicitly preempt state gambling laws for CFTC-registered contracts. The bill stalled, but the New York ruling could revive it. Greed is the feature; the bug is just the trigger. If Congress fixes the bug, the feature—massive profits from prediction markets—remains intact.
Large platforms like Crypto.com and Coinbase have already built geo-fencing infrastructure. They blocked New York users for years. For them, the ruling is just confirmation. They will adapt by adding more state-level restrictions. The cost is manageable at scale. Smaller platforms like Kalshi and Polymarket will struggle, but consolidation is standard in regulated industries. The survivors will be those with the deepest pockets and strongest lobbyists.
Third, the ruling may clarify that event contracts on sports or elections are gambling, but contracts on economic indicators (inflation, GDP) are not. The judge specifically noted that Kalshi's contracts on "elections" were at issue; she left open the possibility that other event contracts might not be gambling. If the CFTC defines certain contracts as "hedging instruments" rather than gambling, states may have less authority to ban them. That's a distinction worth watching.
Takeaway You didn't pay enough attention to the dual sovereignty doctrine. Prediction markets are now a case study in how federal regulatory approval is not a shield against state action. The exploit wasn't in the code; it was in the legal contract. Logic doesn't care about your growth projections; it only cares about the law's structure. The next six months will determine whether prediction markets become a mainstream financial product or a niche, fragmented service. Expect more lawsuits, more geo-blocking, and more uncertainty. If you're investing in this sector, you are betting not on technology, but on the willingness of Congress to override 50 state legislatures. That's a bet I would not take without a very wide margin of safety.
