Consider the quiet arithmetic of a preliminary injunction — a document so dry that most readers will never open it, yet so consequential that it may reshape how tens of millions of Americans place a wager on the future. On September 25, the Sixth Circuit ruled against Kalshi, the federally licensed prediction market that has spent two years insisting its CFTC registration should exempt it from state gaming laws in Ohio and Tennessee. The court rejected not one but two of its central defenses. First, Kalshi failed to demonstrate that its event contracts qualify as "swaps" under the Commodity Exchange Act. Second — and far more damaging — even if they did, federal commodities law does not preempt state gaming statutes. Two doors, closed in a single motion.
I have audited enough smart contracts to recognize the pattern. When a single flawed assumption sits at the base of a system, you do not need to attack the whole structure. You find the assumption, and everything above it collapses. Kalshi's assumption was elegant and, until this autumn, profitable: that a federal license confers a federal monopoly. The Sixth Circuit has now told it otherwise.
This is not, strictly speaking, a Web3 story. Kalshi is not a chain. It is a centralized matching engine — permissioned, KYC-gated, and regulated as a Designated Contract Market, or DCM, under the Commodity Exchange Act. There is no token, no staking, no liquidity mining. Its revenue arrives the old way: transaction fees on contracts that let users speculate on elections, interest rates, weather, and — overwhelmingly — sporting events.
That architecture matters, because it is the mirror image of Polymarket, the on-chain prediction market built on Polygon where users connect a self-custodied wallet and no central operator stands between them and the order book. For years, regulators treated the two as distant cousins — one compliant, one fugitive. That distinction is now collapsing.
What makes the Sixth Circuit's ruling so consequential is not the fine — though Michigan has already published a penalty ceiling of $500,000 per day. It is the structure of the reasoning. The panel ruled that Kalshi's contracts are not provably "swaps" under the CEA, which means the CFTC's exclusive jurisdiction does not automatically attach. And it ruled that, were they swaps, federal law still would not displace state gaming statutes. This "double rejection" gives every state in the circuit two independent paths to victory — and gives Kalshi almost none.
To understand why, you have to understand the technology the court placed at the center of its decision: geofencing. Geofencing is the practice of restricting digital access based on a user's physical location, verified through IP addresses, GPS signals, and billing-address checks. Kalshi argued that excluding users state-by-state was "technically difficult, time-consuming, and expensive." The court's response, delivered without sentiment, is now a line worth memorizing: expensive does not mean impossible.
The panel noted that other companies already run geofencing that satisfies both federal rules and state gaming law simultaneously. In other words, the compliance stack the court demanded already exists. Kalshi's defense failed not because the technology was absent, but because the will to deploy it was. This is the first insight worth absorbing: the court did not invent a new compliance obligation. It simply refused to accept that a costly one was an impossible one.
For a centralized exchange, geofencing is eminently enforceable. Kalshi knows its users' identities, their payment rails, and their device signatures. It can block a Tennessee resident in milliseconds. That is exactly the point. When I spent six hundred hours in 2020 manually auditing the interest-rate models in Aave V2, I learned that the most dangerous bugs are never in the code people admire — they are in the assumptions nobody thought to question. Kalshi's centralization, long marketed as its compliance advantage, has become the lever courts can use to compel it. A permissioned system can be ordered to exclude. A permissionless one, in theory, cannot. But even that asymmetry is about to be tested in ways that should worry every decentralization maximalist.
Because Kalshi was not the only defendant. Kentucky filed suit against Kalshi and Polymarket in the same action. The courts and state attorneys general are quietly classifying both as the same kind of object: a venue where Americans wager on outcomes. The architecture — on-chain versus off-chain, custodial versus self-custodial — is being flattened into irrelevance at the judicial level. When I helped draft the open-source charter for the DAO Guilds movement, this was the fault line I kept returning to: legal systems do not care about your decentralization diagram. They care about who holds the money and who can be served with a summons.
Now consider the economics, because they expose how thin Kalshi's underlying position really is. In 2025, sporting contracts account for more than 90 percent of its trading volume and roughly 95 percent of its revenue. This is not an information market in the intellectual sense the phrase implies. It is a sportsbook wearing the language of collective intelligence. And it is a textbook example of single-point dependency — the kind of revenue concentration I would flag in red ink if I found it on a project's token dashboard.
The second insight is sharper still: 69 percent of Kalshi's retail sports demand originates in states that lack legal online sports betting. In those jurisdictions, Kalshi is not competing on product quality — it is competing on mere availability. It is the only door in town. The substitution rate inside licensed-betting states, where real competition exists, is a mere 2 to 4 percent. Read that again. Kalshi's growth is not a testament to superior technology or superior forecasting. It is a testament to a regulatory vacuum. Remove the vacuum, and the growth evaporates.
The financial motive behind the state lawsuits is equally clear. Licensed sportsbooks paid more than $3.2 billion in state taxes during fiscal year 2025. Kalshi, operating outside those rules, paid a rounding error. That asymmetry — $3.2 billion against near nothing — is not a legal footnote. It is the engine driving fifty attorneys general toward the same conclusion. Fiscal hunger does not tire. States will not simply walk away.
The geographic exposure concentrates the risk to a razor's edge. California and Texas together account for roughly 44 percent of Kalshi's modeled demand. California sits in the Ninth Circuit, which on August 28 already ruled against Kalshi's preemption claim in a Nevada matter. Texas sits in the Fifth Circuit, which has not yet decided. If California is already hostile and Texas is the pivot, then the entire domestic future of regulated prediction markets may now rest on a single appellate court in New Orleans. I have watched enough governance votes and audit reports to know that when a system's viability depends on one undecided variable, that is not a business model. That is a coin flip with a fee schedule.
Michigan has already shown what enforcement looks like. Its regulators published a blocking order with a daily penalty ceiling of $500,000, and Kalshi complied. This is the template — not a theoretical threat but a demonstrated, replicable mechanism. Any state can copy it tomorrow. The penalty is not designed to extract revenue; it is designed to force behavioral change before the litigation concludes. It works. And it means Kalshi may be compelled to geofence itself into irrelevance long before any final judgment is rendered. The courtroom clock and the business clock have started running at different speeds, and the business clock is losing.
Here is where a technical observer must go beyond the legal filings. Anyone who can read a sanctions list can see that the DCM's national order-matching obligation was reinterpreted in this ruling: the court read the federal requirement as applying only to markets the exchange itself chooses to list. That quietly opens a door for "product-level plus geography-level" segmented operation. A DCM may now legally run one product in Ohio and withhold it in Tennessee. That is a subtle but real precedent — and it is the kind of architectural detail that will matter more than any headline.
The downstream effects follow the same logic as any dependency graph. Regulatory technology — geolocation providers, address-verification services, proof-of-location infrastructure — stands to benefit as every prediction venue is forced to prove where its users physically sit. DeFi and on-chain prediction ecosystems face the opposite pressure: if a centralized exchange can be compelled to geofence, the "we cannot block users" defense becomes far harder to sustain in court. And the licensed gaming operators, already moving in through products like DraftKings Predictions, are the quiet winners. The transition now underway is simple to state: arbitrage players exit, licensed players enter.
There is a convenient objection — that geofencing is porous, that VPNs defeat it, that location proofs are spoofable. This is true and irrelevant. Courts do not require perfect enforcement; they require demonstrable good faith. The very existence of a circumvention tool does not excuse an operator from deploying the wall.
Now the counter-intuitive part — the blind spot almost everyone covering this story is missing.
The instinctive reading is that Kalshi's defeat is a gift to the on-chain camp. If the centralized, compliant player is being squeezed by state law, surely the decentralized alternative looks vindicated. I think this reading is exactly backwards.
Kalshi's core defense was that state-by-state compliance was technologically infeasible. The court demolished that defense by pointing to existing tools. But that demolition does not stop at Kalshi's front door. It sets the evidentiary bar for everyone. When Polymarket eventually attempts its American return through a CFTC-registered entity — as it has signaled — it will inherit a legal record in which a far more pliable, centralized competitor was told that geofencing is both necessary and achievable. "We cannot block wallet addresses reliably" will be heard by judges who have just been shown that reliability is a matter of spending money, not of physics.
The uncomfortable truth is that Kalshi's loss is worse news for decentralization than for Kalshi itself. Code is law, but ethics is soul — and the courts have just demonstrated that they will treat the soul of an on-chain protocol and the body of a centralized exchange as morally equivalent when the public interest is at stake. The regulatory arbitrage window that both Kalshi and Polymarket have quietly occupied for years is closing, not because the technology failed, but because the state has decided it will not tolerate being routed around.
What the industry's marketing refuses to concede is this: the "information market" narrative was never the real product. The real product was access — access to outcomes that citizens in certain jurisdictions were legally forbidden to wager on. That is not a prediction market. That is a regulatory loophole with a polished interface, and loopholes, by definition, close.
So watch the Fifth Circuit in Texas, because it now controls the fate of roughly 44 percent of the demand that keeps this sector breathing. If it rules for Kalshi, we may see a genuine circuit split — and a case that climbs toward the Supreme Court. If it rules against, the arbitrage model dies quietly, state by state, penalty by penalty, until there is nothing left to appeal.
Transparency isn't the oxygen of trust; it is the architecture that makes trust possible at all. The prediction market industry is about to learn whether it was ever building trust — or merely renting access from a landlord who has finally come to collect.