The Broken Handshake: Kalshi's Montana Complaint and the Fraying Truce Between States and the CFTC

On April 16, 2025, lawyers for KalshiEX LLC and lawyers for the Montana Department of Justice's Gambling Control Division reached something rare in the current prediction-market wars: a written understanding. Montana would hold its fire. Kalshi would keep operating. Both sides would wait to see what Chief Judge Gordon did in Las Vegas, and whatever happened in the Ninth Circuit would shape what happened next. It was a lawyer's truce, the kind that usually holds because neither side benefits from breaking it in public.

It held for 355 days.

On April 6, 2026, Montana's Gambling Control Division sent Kalshi a second cease-and-desist letter. The division's chief legal counsel had, in the intervening months, taken a second look at the truce and decided it had been contingent all along: contingent on Judge Gordon's preliminary injunction remaining in effect, contingent on a particular posture in the Ninth Circuit, contingent on a reading of the April 16 correspondence that Kalshi's lawyers say was never on the table. Six days later, Kalshi sued. The complaint hit the docket of the U.S. District Court for the District of Montana on Sunday, April 12, 2026, naming Attorney General Austin Knudsen and several Gambling Control Division officials as defendants, and asking for a preliminary injunction, a permanent injunction, and declaratory relief.

The Montana filing is, on one level, a routine addition to Kalshi's expanding legal map. The company has been in federal court against Nevada, New Jersey, Maryland, Arizona, Washington, and several other state regulators. The CFTC, under Chair Mike Selig, has made this its signature litigation project, filing its own offensive suits against Illinois, Connecticut, and Arizona on April 2, 2026. Montana, though, is different. It is the first state to openly walk away from a negotiated non-enforcement posture, and the complaint is a document about trust decaying in real time. Every founder building a product that sits inside two regulatory frameworks at once should read it twice.

What actually happened

Kalshi's April 12 complaint attaches two cease-and-desist letters as exhibits. The first, dated March 2025, came early in the state-level pressure campaign and read like most of the others: Montana characterized Kalshi's sports event contracts as unlicensed sports wagering under state law and demanded the company stop offering them in the state. Kalshi's outside counsel met with division representatives in Helena soon after. According to the complaint, the parties reached the following understanding, which Montana's side confirmed in writing on April 16, 2025: the state would defer enforcement pending the outcome of the Nevada proceedings before Chief Judge Andrew Gordon and any appeal to the Ninth Circuit, and would give Kalshi reasonable notice before commencing any civil or criminal action.

That understanding then met the facts on the ground. Judge Gordon granted Kalshi a preliminary injunction in Nevada in spring 2025, and in November 2025 reversed himself and dissolved it. The Ninth Circuit hears consolidated oral argument on April 16, 2026, in appeals from Kalshi, Robinhood, and Crypto.com. Nothing in the Ninth Circuit has resolved. Montana's April 6 position is that the deferral clock ran out when Judge Gordon dissolved his injunction. Kalshi's position is that the deferral was pegged to a final outcome on appeal, not to a volatile interim posture, and that Montana's move is both a breach of the April 16, 2025 correspondence and an act of federal preemption by a covered state officer.

Kalshi's legal theory in the Montana complaint is the one its counsel has now made familiar: as a CFTC-designated contract market, Kalshi is regulated under the Commodity Exchange Act, its sports event contracts are "swaps" within the meaning of 7 U.S.C. § 1a(47), and state enforcement against such products is barred by field preemption, conflict preemption, and, in the company's view, express preemption as well. The complaint asks the District of Montana to enjoin the state preliminarily and then permanently, and to declare the April 6 cease-and-desist void.

Which beat this is, and why

This event qualifies on two beats at once. It is an enforcement action (beat three): a state regulator issuing a cease-and-desist against a named company, and that company responding with a federal injunction action in which the state will either stand down or litigate. It is also a regulatory development (beat two): Montana's move reshapes the posture of a state-federal dispute that is now at the circuit court level in three different places, and the terms on which state regulators accept or reject CFTC-preemption arguments matter for every vendor in this space. It also qualifies, secondarily, on beat four (game mechanics) because the underlying dispute is over a specific contract form: single-game sports event contracts denominated in dollar-per-contract payouts.

Our recent coverage of the broader Kalshi litigation posture already sits in the series (see "The Kalshi Gambit," April 13, 2026, and "The End of Category Lines," April 13, 2026). This is explicitly an update piece, narrowly focused on what the Montana filing and the Arizona restraining order, issued April 10, 2026, teach us about how this fight is now being fought at the tactical level.

A. The case at the center

KalshiEX LLC v. Knudsen, et al., filed April 12, 2026, U.S. District Court for the District of Montana. Docket number was not yet public at the time of our review; verify on PACER or CourtListener before citing in filings. The complaint seeks declaratory and injunctive relief against Montana Attorney General Austin Knudsen in his official capacity and several officials of the Gambling Control Division, and attaches the March 2025 and April 6, 2026 cease-and-desist letters as exhibits, along with the April 16, 2025 correspondence on which Kalshi's breach-of-understanding argument turns.

The factual record to watch is twofold. First, how the court characterizes the April 16, 2025 correspondence: as a binding agreement, a representation, or a courtesy the state remained free to rescind. Montana's best argument is likely that the letter was an expression of prosecutorial discretion contingent on the legal landscape remaining in a certain posture, and that the landscape shifted materially when Judge Gordon dissolved his own injunction in Las Vegas. Kalshi's best argument is that the letter names the Ninth Circuit appeal as the trigger, not the trial-court preliminary injunction, and that Montana's unilateral reinterpretation is both a breach of the representation and textbook field preemption. Second, the procedural posture: Kalshi is likely to seek a temporary restraining order first, modeling the ask on what the District of Arizona just granted.

That Arizona TRO, issued by Judge Susan Brnovich on April 10, 2026 and running through April 24, blocks Arizona Attorney General Kris Mayes from continuing her first-in-the-nation criminal case against Kalshi while the federal preemption question is litigated. The criminal complaint, filed in Maricopa County Superior Court on March 17, 2026, charged Kalshi with operating an unlicensed gambling business under Arizona Revised Statutes § 13-3303 and with accepting wagers on elections. The CFTC and the Department of Justice appeared as movants and asked the federal court to enjoin the state criminal case on preemption grounds. The district court agreed, at least through April 24, and Mayes's office has said it will not move the Maricopa County case forward while the TRO is in place. The Arizona case gave Kalshi and the CFTC a template. Expect the Montana complaint to follow the same choreography.

B. The historical precedent that frames this

Every one of these prediction-market preemption fights runs through Murphy v. NCAA, 138 S. Ct. 1461 (2018). Murphy struck down the Professional and Amateur Sports Protection Act under the anti-commandeering doctrine, returned sports betting regulation to the states, and has been, for eight years, the cornerstone of state authority over sportsbook operators. Every state that issued a cease-and-desist to Kalshi did so on the implicit theory that Murphy put this police power squarely in state hands. The theory Kalshi is advancing inverts Murphy: yes, states got sports betting back, but only sports betting. Derivatives on sporting outcomes, traded on CFTC-designated contract markets, were never inside PASPA's domain and are now inside the CEA's.

The Third Circuit's April 6, 2026 opinion in KalshiEX LLC v. Flaherty, No. 25-1922, accepted that inversion on a preliminary-injunction record. Judge David Porter, writing for a 2-1 majority joined by Chief Judge Michael Chagares, held that the CEA preempts state gambling laws as applied to sports event contracts traded on a CFTC-registered DCM. He walked through the 1974 amendments that created the CFTC, read New Jersey's enforcement effort as exactly the "patchwork" Congress wanted to displace, and held that Kalshi's contracts are "swaps" because their value is plainly associated with financial, economic, and commercial consequences (advertisers, sponsors, television rights, gambling markets themselves). Judge Jane Roth's dissent leaned on the historical presumption against preemption in areas of traditional state police power, specifically vice regulation, and warned that DCM trading is too narrow a subfield to sustain field preemption. She has the better of the 1936-to-1974 historical argument; the majority has the better of the textual argument. A genuine circuit split is now emerging between the Third Circuit on one hand and the Ninth Circuit posture in the Nevada litigation on the other, with the Fourth Circuit appeal from Maryland still pending.

Murphy is the one the panels cite. But the case that actually drives the outcome here is more likely to be Altria Group, Inc. v. Good, 555 U.S. 70 (2008), on the scope of express preemption, and Arizona v. United States, 567 U.S. 387 (2012), on field and conflict preemption doctrine. How the Supreme Court, if it takes one of these, reconciles traditional state police power with the CEA's comprehensive scheme will determine not only Kalshi's business model but the scope of dozens of state licensing regimes.

C. Live rulemaking and legislation to watch

The CFTC's own event-contract rulemaking is the single most important regulatory instrument in this space. CFTC RIN 3038-AF20 (the event-contract rule-proposal being advanced by Chair Selig) has a comment period that closes April 30, 2026. Whatever the agency finalizes will itself become a battleground. A narrow rule that defines "gaming" contracts restrictively will strengthen Kalshi's preemption story. A broad rule, or one that gives the CFTC more discretion to disapprove individual contracts under 17 C.F.R. § 40.11, will weaken it, because the CFTC will then be, in effect, the one saying that some of this is gambling after all. Rule 40.11(a)(1), which already bars DCMs from listing contracts involving certain enumerated categories, was cited by the Third Circuit dissent as evidence that the CFTC itself treats gaming contracts differently, which the dissent viewed as a fatal crack in the conflict-preemption theory.

On the state side, at least eleven states have introduced 2026 legislation that would ban, license, or tax prediction markets. Watch especially:

- California AB 831 (Valencia), pending, the vendor-liability expansion we covered last week, which sweeps in payment processors and service providers for sweepstakes and, if amended, could be framed to reach prediction-market vendors.
- New York S 5935-A, introduced in February, which would prohibit a licensed sportsbook from also operating a prediction market under a different regulatory theory in the same state.
- Illinois HB 4154, which would explicitly criminalize offering event contracts on sports outcomes to Illinois residents.
- Connecticut's DCP cease-and-desist and the pending state legislation backed by Governor Lamont that would ban prediction markets from being used by or advertised to consumers under 21.

Federal activity beyond the CFTC rulemaking: the Department of Justice has intervened twice now in support of Kalshi, once in Arizona and once in support of the CFTC's three offensive suits. The DOJ's position is an unusual one, taking the federal preemption argument further than the CFTC has taken it in its own filings. Treasury and FinCEN have not yet weighed in, but if prediction markets start representing serious cross-border dollar flow, expect FinCEN guidance on bank secrecy act applicability to DCM-traded event contracts within six to twelve months. Tribal governments are also a factor. The Blue Lake Rancheria case (Authorities for Blue Lake Rancheria, et al. v. Kalshi, Inc., et al., No. 25-7504) frames prediction markets as a potential IGRA violation and as a threat to tribal exclusivity compacts. That suit has its own calendar.

D. The real points of contention

Three fights are happening underneath the preemption caption.

The first is whether a single-game sports event contract is a swap at all. Kalshi's lawyers lean on the 2010 Dodd-Frank definition: any agreement whose payoff is contingent on an event associated with a financial, economic, or commercial consequence. The Third Circuit majority accepted this expansively. Judge Roth's dissent and the Nevada position accept it narrowly and argue that a bet on the Kansas City Chiefs is a bet on the Kansas City Chiefs, regardless of how it is documented. The Supreme Court has never addressed the outer boundary of "swap" under § 1a(47). It will.

The second is whether state gambling statutes fall within the CEA's savings clauses. 7 U.S.C. § 2(a)(1)(A) and related provisions preserve some state authority. How much? The Third Circuit majority read them narrowly. The Ninth Circuit, so far, has read them more generously to the states. Which reading wins at One First Street is a coin flip.

The third, the one the Montana complaint sharpens, is whether state AGs and state gaming divisions can withdraw from individualized non-enforcement representations without notice. This is not a preemption question, it is an estoppel and due-process question, and it will get lost in the preemption briefing unless Kalshi's Montana counsel pushes it. But it matters for every operator and vendor that has negotiated a similar informal posture with a state. If Montana succeeds in reading its April 16, 2025 letter as conditional and revocable at will, every one of those operator-state understandings becomes worthless paper, and the industry loses a tool it has relied on for two decades.

E. Our analysis: what operators, founders, and vendors should actually do

Three audiences read this differently, and they should act on it differently.

For existing licensed operators running conventional sportsbooks, the Montana and Arizona events change nothing operational but change the competitive calculus. FanDuel and DraftKings have hedged by quietly launching or planning prediction-market surfaces of their own. The question is no longer whether the prediction market exists in your jurisdiction but whether you can afford not to compete with it on your own regulated rails. Assume, for planning purposes over the next 12 months, that Kalshi continues operating in Montana, Pennsylvania, New Jersey, Delaware, and any state in the Third Circuit's domain, and that Nevada and Washington stay contested. Budget for a Supreme Court ruling in the October 2026 or October 2027 term. Run two product roadmaps.

For founders building adjacent products, the Kalshi map is now a template. If your product is a contract, a token, or an instrument whose payoff is contingent on a real-world event with economic consequences, and you can register as a DCM or list on one, you have a federal-preemption argument available to you. It will not be cheap: expect $8 to $20 million on litigation in the first 36 months. Make sure your investors understand what they are funding. If your product is closer to a sportsbook, with the house on the other side of the trade, the preemption argument is much weaker and the operational license path is the only real path. Do not try to reverse-engineer a DCM on top of a house-banked product. That is the path to an Arizona-style criminal complaint.

For vendors and service providers, especially compliance, geolocation, KYC, payments, and data, Montana is a signal to formalize everything. The regulators writing cease-and-desists to Kalshi are also writing them to vendors. Michigan sent 45 such letters this spring. California AB 831 expands vendor liability explicitly. The single most important action in the next 90 days: reduce your legal exposure to contested states. That means jurisdictional carveouts in your MSA, an express indemnity from the operator covering state enforcement against you, and a documented compliance posture.

Decision tree: (i) licensed operator, licensed state: continue as is, get the indemnity in writing; (ii) DCM-registered prediction market, contested state: reduce your footprint to service delivery from outside that state where possible, get written acknowledgment that the operator, not you, is the regulated party; (iii) DCM-registered prediction market, state with live criminal enforcement (Arizona, Washington): consider suspending delivery to residents of that state until the CFTC's preemption theory is resolved in the relevant circuit.

The Founder's Playbook

The thing that breaks in the Montana complaint is not a legal theory, it is a handshake. For more than two decades, the single most reliable tool any gaming operator has had for managing multi-state exposure has been the quiet prosecutorial-discretion letter, often unwritten, usually from a state gaming enforcement lawyer to operator counsel, saying in effect: we have concerns, but we are not going to move on you while the larger legal question works itself out. That tool has lubricated everything from DFS operations in the years after the 2015 New York AG action to early sweepstakes casinos to daily fantasy sports in the hinterlands. When the Montana Gambling Control Division decided, in April 2026, that its April 2025 representation was contingent on a fact pattern that had since changed, it did something most state regulators have been unwilling to do. It walked away from the handshake in public.

Two lessons for founders. First, papers matter more than they used to. If you have an informal non-enforcement understanding with a state regulator, document it with specificity, identify the triggering events for its end, and have outside counsel send a confirmatory letter the state can correct if its view differs. The Montana correspondence was exactly the kind of document that should have been airtight; its ambiguity made the April 6 retreat possible. Second, diversify your legal posture. Operators with concentrated exposure in one or two friendly states in 2025 now face a patchwork in 2026, and the winners will have built, early, the litigation capacity and product flexibility to operate where rules differ across state lines for at least 24 months.

Watch list

Developments likely in the next two to six weeks:

- A TRO decision in Kalshi v. Knudsen, likely within 14 to 21 days, probably favorable to Kalshi given the Arizona and Third Circuit patterns.
- Ninth Circuit oral argument on April 16, 2026 in the consolidated Kalshi/Robinhood/Crypto.com appeals against Nevada. The panel (R. Nelson, Bade, K. Lee) is unusually friendly to the federal-preemption theory.
- Close of comments on CFTC RIN 3038-AF20 on April 30, 2026.
- A possible en banc petition from New Jersey in the Third Circuit, due in late May 2026.
- A second round of state cease-and-desists from jurisdictions that have so far stayed out. Texas, Florida, Ohio, and Massachusetts are the most likely candidates.
- A follow-on Kalshi filing in Washington or Oregon.

The map is moving fast. The handshake is, for now, broken. The next move is Montana's.