After a college hockey market settled in a way nobody with a ticket expected, one trader asked the question this article exists to answer: "Kalshi changed the rules AFTER market close - is this illegal?" The direct answer: Kalshi does hold written powers to modify contracts after trading opens, and exercising them through the filed federal process is lawful, because those powers sit in a rulebook filed with the CFTC, the federal regulator that oversees the exchange. Whether one specific settlement used those powers the way the law requires is a separate question, and a pending lawsuit is testing exactly that. But there is a catch that cuts both ways. Ordinary rule changes have to be filed with the CFTC before they take effect, and emergency actions carry their own notice obligations, so a silent midnight rewrite is not what the system permits. What burned traders in the most famous cases was usually something different: fine print that existed, or discretion the rulebook already granted, surfacing only at settlement.
I will walk through the three rulebook powers that let Kalshi change things, the federal filing clock that constrains them, and the $54 million market where users allege the line was crossed. In my read of every public blow-up in this class, one clause keeps deciding the fight, and I will get to it, because it is the one you should reread before your next trade.
The Quick Answer
Yes, Kalshi can modify contracts after you enter a position: Rule 7.2 lets it swap data sources and move expiration dates, Rule 2.8 grants broad emergency powers, and Rule 6.3(c) gives it "sole discretion" to interpret a contract's terms at settlement. The check on those powers is process, not prohibition: routine rule amendments must be filed with the CFTC 10 business days before implementation, and emergency actions must be reported within 24 hours and announced to traders. The three powers, the filing clock, and the death-carveout lawsuit now testing the limits are below.
Power One: Rule 7.2 Lets Kalshi Modify A Contract Mid-Flight
Start with the rulebook itself, because every claim here should trace to a document you can open. The KalshiEX rulebook (v1.18, fetched August 13, 2026) is filed publicly with the CFTC, and Rule 7.2, titled Contract Modifications, says Kalshi "retains the authority to designate a new Source Agency and Underlying for that Contract and to change any associated contract specifications after the first day of trading" when something materially impacts the reliability of the data source a market settles on. It can also adjust a market's expiration date when "the rescheduling or cancellation of an event" or delayed data would prevent an accurate settlement, and pull expiration earlier when the outcome a market asks about has already happened.
"Change any associated contract specifications after the first day of trading" is an explicit, regulator-filed power to alter a live market. The constraint attached to it is disclosure, not permission. Every branch of the rule carries the same duty, that a source swap "would announce any such decision on its website" and an expiration change "will be announced on its website." So the honest answer to "can they change it after I bought in" is yes, and they told you so in advance, in a document most traders never open. That asymmetry is where the angriest settlement threads are born: the powers are in writing, and most readers stopped at the market title. That is why the sharpest habit in how Kalshi settlement works is reading the full rules tab before you click buy.
Power Two: Rule 2.8 Emergency Powers Can Rewrite Terms In Real Time
Rule 7.2 handles broken data sources. Kalshi emergency rules, housed in Rule 2.8, handle everything else. In a declared emergency, the board or two members of management, with CEO approval, "may, without giving prior notice to, or securing prior approval from the Commission, adopt a temporary emergency rule." The definition of emergency is broad: market manipulation, government action, system failures, and a catch-all for "any other unusual, unforeseeable, and adverse circumstance" threatening fair and orderly trading.
The listed emergency actions are the full toolbox. Rule 2.8(d) permits, among other things, "Cancellation of a Contract and the return of any funds paid to enter Trades on the Contract," suspension of trading, moving expiration dates, and, most relevant to our question, "Changing a Contract's terms and conditions and/or specifications." The rulebook is blunt about what this means for you, stating in 2.8(b) that "the chance of an emergency is one of the risks that Participants should consider when deciding whether to trade on the Platform."
Because Rule 2.8 is the fastest-moving power, it also carries the tightest transparency leash, and this is the part the angry threads usually miss. Under 2.8(e) and (f), Kalshi must notify the CFTC no more than 24 hours after implementing an emergency action, publish a notice on its website, notify members through the API and by email, and document its reasoning. An emergency change you never hear about is not how the rule is written to operate.
The Clause That Decides Most Disputes: Sole Discretion At Settlement
Here is the clause I promised, the one that in practice keeps resolving the "they changed the rules" feeling: Rule 6.3(c). It states that "Kalshi has sole discretion to interpret a Contract's Terms and Conditions," and that when a contract expires without a clearly determinable outcome, Kalshi "may use the last traded price of the Contract to determine the payout." The rulebook's own example: a contract that last traded at 10¢ for the long side, an implied 10% chance, pays longs 10¢ and shorts 90¢ (the short side's 90%), instead of the $1.00 a YES resolution would pay.
Notice what this clause is not. It is not an amendment power at all; nothing in the contract changes. It is an interpretation power that activates exactly when reality lands in a spot the market title never contemplated, and that timing is why it feels like a post-close rewrite from the outside. A college hockey game that ends tied and goes to a shootout is a perfect specimen: sport-specific contract terms define result scope in advance, the way Kalshi's soccer terms spell out whether a market covers regulation time only or includes a penalty shootout, and the way its football contract terms pre-commit to a separate kind of fine print, a 48-hour postponement clock with a "last fair market price" fallback. When a scope clause and the scoreboard diverge, the clause wins, the market settles on the clause's reading, and the trader who never read it experiences a rule change even though the words sat in the terms since the market opened. The same last-traded-price mechanics that govern voided situations show up across the exchange, including in how parlay leg voids work.
If 6.3(c)'s last traded price is unavailable or unfair, an Outcome Review Committee makes what the rule calls a binding determination, "final and not subject to review." The rulebook houses a related mechanism in Rule 7.1, the Market Outcome Review Process, which Kalshi can open before settlement when something material hits the underlying data; that process runs on its own 24-hour clock and its determinations are likewise final. Full anatomy of those criteria lives in Kalshi's resolution criteria, explained.
The Federal Check: A 10-Day Filing Clock, Not A Blank Check
None of these powers float free, because Kalshi is a designated contract market under CFTC oversight, which separates it from an offshore book in ways that matter here (the full comparison is in Kalshi vs. sportsbooks). Under federal regulation 17 CFR 40.6 (fetched August 13, 2026), a rule change must be submitted to the CFTC "not later than the open of business on the business day that is 10 business days prior to the registered entity's implementation," with the amended text marked, notice posted on the exchange's website, and a certification that "the rule complies with the Act and the Commission's regulations thereunder." The CFTC can stay the change during that window.
Emergency rules run on the inverted clock you already saw in Rule 2.8: they may be filed "at the earliest possible time after implementation, but in no event more than twenty-four hours after implementation." Put the two clocks together and a Kalshi rule change has exactly two legal speeds. Fast, when it declares an emergency and must answer for it within a day; and slow, through a filing the public can read 10 business days ahead. So can Kalshi change rules after market close? Through one of those two doors, yes, and through no other. What the regime does not contemplate is a rule that changes with no filing and no notice at all.
The Death Carveout Case: Where Users Allege The Line Was Crossed
Every mechanism above is now being stress-tested in one lawsuit, the Kalshi death carveout case. Kalshi's "Ali Khamenei out as Supreme Leader?" market (ticker KXKHAMENEIOUT) opened January 8, 2026 and drew more than $54.5 million in volume. When Khamenei was killed in airstrikes on February 28, 2026, YES holders expected their contracts to resolve YES at $1.00. Instead, Kalshi settled the market at the last traded price before the news, 9¢, an implied 9% chance, citing fine print that payouts would be determined "based upon the last traded price prior to death."
A class action, Risch, et al. v. KalshiEX LLC, No. 2:26-cv-02390 (C.D. Cal., filed March 5, 2026), now claims that carveout was the wrong kind of surprise. Plaintiffs allege the death-carveout disclaimer was added only after military strikes began, that trading continued through February 28 while the exchange knew how it would settle, and that an ordinary reader of "out as Supreme Leader" would understand death to qualify. The complaint calls the episode "the poster child of unfair competition, deceptive corporate behavior, and consumer fraud." Those are allegations in a pending case, not established facts, and the timing question they raise is precisely the one this article's mechanisms frame: a pre-existing carveout applied through settlement discretion is business as usual, while a term added after close would be a different animal entirely. Which of those happened here is what the court will decide.
Kalshi's side of the ledger is real too, and it responded in the currency exchanges actually control. Per Casino.org's reporting, the company conceded the market's rules were "grammatically ambiguous," said "similar contracts in the future will be better explained," and reimbursed all trading fees and net losses on the market "out of pocket, so no trader ended net negative." CEO Tarek Mansour has said Kalshi does not list "markets directly tied to death" and designs rules "to prevent people from profiting from death." Refunding losses on a disputed settlement is not the same as paying winners, though, and coverage of the filing notes that gap is the heart of the suit.
Worked Example: What A 9¢ Settlement Does To A YES Ticket
Run the Khamenei settlement math on a clean position, using the market's real prices. Say you put $60 on YES at 6¢, an implied 6% chance, which buys 1,000 contracts.
| Scenario | Rule applied | Your 1,000 contracts pay | Net result |
|---|---|---|---|
| Market Resolves YES At $1.00 | Rule 6.3(a) standard binary settlement | $1,000.00 | +$940.00 |
| Settles At Last Traded Price, 9¢ | Death carveout via 6.3(c)-style discretion | $90.00 | +$30.00 |
| Contract Canceled In An Emergency | Rule 2.8(d) cancellation clause: funds returned | $60.00 back | $0.00 |
The row worth staring at is the middle one, because it shows how violently the two readings of one sentence diverge. The same ticket is worth $1,000.00 under the headline reading and $90.00 under the fine-print reading, a 91% haircut from the payout the buyer believed was locked. And notice the strangeness of that middle row: because the 9¢ settlement sits above the 6¢ (6%) entry, the position still shows a $30.00 paper profit, even as its holder feels robbed of a $940.00 win. One trader told Casino.org he had put $303.90 into YES at 6% implied odds, expecting roughly $4,588, and was paid $91.76 before Kalshi's later refunds. That spread between expected and received, not the direction of the news, is what filled the settlement threads, and it all hinged on which clause governed. Now run the same 6.3(c) mechanics on the other side of the book: a NO position bought at 94¢, a 94% implied chance, would be paid 91¢ per contract under a 9¢ carveout settlement, a near-scratch 3¢ loss on a position the headline reading would have zeroed out entirely.
What This Means Before Your Next Trade
The college hockey question deserves its closing answer in plain terms. Was it illegal? Changing settlement terms after close, if that is what actually happened, is the exact theory a federal class action is testing right now, so no one should tell you the question is settled either way. Applying an already-filed clause you did not know about is, by contrast, the system working as designed, uncomfortable as that is. The practical defenses follow directly from the mechanisms:
- Read The Market Rules Tab, Not Just The Title, Before You Trade. The title is marketing; the Payout Criterion is the contract. Result scopes (overtime, shootouts, "prior to death") live there, and under 6.3(c) the clause outranks your reading of the headline.
- Price In The Backstop. When an outcome cannot be cleanly determined, the default is settlement at the last traded price, not a void or a $1.00 payout. If you would not accept that risk at 6¢ (6%), at a coin-flip 50¢ (50%), or at a favorite's 60¢ (60%), size the position accordingly.
- Watch The Exchange's Own Channels, And Know The Dispute Path When They Fail You. Rule 7.2 announcements and Rule 2.8 emergency notices land on Kalshi's website, API, and email; a change you cannot find in those channels is worth escalating through the formal complaint route in how Kalshi settlement disputes work. Fees on the Khamenei market were ultimately refunded, and the fee mechanics are in our Kalshi fees breakdown.
Fine print risk is one more input to price, exactly like vig or liquidity, and traders who price it are the ones still holding an edge when a market gets weird. If you want the value side of that discipline done for you daily, our free expert picks show where our analysts see real edges across books and exchanges, with zero of your bankroll required to look.
FAQ: Kalshi Rule Changes
Can Kalshi change rules after market close? Routine amendments cannot legally take effect without a CFTC filing 10 business days in advance, and emergency changes must be reported to the CFTC within 24 hours and announced to traders. What Kalshi can do at close is interpret existing terms under Rule 6.3(c), including settling an unclear market at its last traded price.
Can Kalshi cancel a market and keep my money? No rule permits keeping your funds on a cancellation. Rule 2.8(d) pairs contract cancellation with "the return of any funds paid to enter Trades on the Contract." The scenario traders actually face is different: a settlement at the last traded price, where a 9¢ close pays 9¢, not a refund.
Did Kalshi break the law in the Khamenei market? That is unresolved. A class action filed March 5, 2026 in the Central District of California alleges the death carveout was disclosed too late; Kalshi enforced the fine print as written while conceding it was "grammatically ambiguous," and it refunded fees and net losses. Courts, not headlines, will settle it.
Where do I find a market's real rules? Every Kalshi market has a rules section listing its Payout Criterion, source agency, expiration, and any carveouts, and sports markets carry sport-specific contract terms documents. Reading that section is the single highest-value 90 seconds in prediction market trading.
