Legally, no — prediction market contracts are federally regulated derivatives, not gambling, and that is the basis on which Kalshi and Novig operate nationwide. Practically, the answer is less tidy, and the fight over which framing wins is the defining regulatory question in this category right now.
Here is what each side actually argues, where the law currently stands, and what it means for you as someone deciding whether to trade.
The legal position
Kalshi and Novig hold designated contract market licences from the Commodity Futures Trading Commission. A DCM is a federal exchange registration — the same category CME and ICE hold. Under that framework, an event contract is a derivative: a financial instrument whose value depends on an outcome, traded on a regulated exchange, subject to federal market rules.
That is not a loophole. It is a licence issued after review, and it pre-empts state gaming law by design. The whole point of a federal exchange registration is that it operates nationally rather than requiring approval in fifty jurisdictions.
CFTC Chairman Selig has publicly supported federal pre-emption on this question, which matters because the regulator's own view carries weight in the litigation.
Why states disagree
Several states see the same product and reach the opposite conclusion.
Their argument runs roughly: whatever the instrument is called, a contract on which team wins Sunday is functionally a bet. The consumer is a retail customer risking money on a sporting outcome, the economics are identical to a wager, and state gaming regulators exist precisely to oversee that activity — including the consumer protections, tax revenue and responsible-gambling requirements that come with it.
This is not a frivolous position. A federal derivatives licence was not designed with sports betting in mind, and states that built regulated sportsbook markets see a competitor operating outside the rules those books must follow.
Where the fight stands
Eleven states introduced prediction market legislation during 2026. The courts have split.
Tennessee ruled in Kalshi's favour. Nevada, Massachusetts, Maryland and Ohio ruled against. Arizona went further, filing criminal charges against Kalshi in March 2026.
In Congress, the Schiff-Curtis bill — titled the "Prediction Markets Are Gambling Act" and introduced in March 2026 — would settle the question in the other direction by statute.
None of this is resolved. Anyone telling you the legal status is settled, in either direction, is overstating it. Our guide to prediction market legality tracks the state-by-state position as it moves.
What genuinely differs from gambling
Setting the law aside, there are real structural differences worth understanding.
No house. A sportsbook takes the other side of your bet and profits when you lose. A prediction market matches you against another user and profits either way. That removes the conflict of interest that sits at the centre of the sportsbook model — and the reason sportsbooks limit or ban winning customers, which is the origin story behind Novig.
Lower margin. A two-way sportsbook line typically holds 4–5%. A liquid prediction market often holds under 1%. You can measure this yourself on any market with our vig calculator.
You can sell. Prediction market positions can be exited before the event resolves. A placed bet cannot. That makes a contract a tradeable asset rather than a locked wager, which is the most substantive argument for the derivative framing.
Price discovery. Aggregated forecasts from these markets are genuinely informative — Polymarket has posted Brier scores around 0.09, roughly 94% accuracy. Nobody cites a sportsbook's line as a research source, though the honest counter is that a sportsbook line is also a forecast and is also accurate.
What does not differ
You can lose everything you stake. A contract settling at $0 is a total loss on that position, and most retail participants lose over time.
It is addictive in the same way. Fast resolution, frequent action, and the same reward loops. A regulated exchange with clean order books produces the same behaviour as a betting app if you engage with it the same way. Kalshi and Novig both publish responsible trading frameworks for exactly this reason, and Novig requires users to be 21+.
The sports markets are the sports markets. More than 80% of category volume is sports. Calling it price discovery does not change what most of the activity is.
What it means for your taxes
The classification fight has a practical consequence that arrives every April.
The IRS has published no guidance specific to prediction markets. Three treatments are in common use: ordinary income, which is what most filers use and the conservative choice; gambling, which is unfavourable because losses are capped — and the OBBBA 90% gambling loss cap starts with tax year 2026; and Section 1256, which is aggressive but arguable for CFTC-regulated contracts.
Notice that the gambling-versus-derivative question is the same question, arriving through a different door. Our prediction market tax guide works through each treatment, and the take-home calculator shows what each does to a given year's profit.
How to think about it for yourself
The legal classification determines what platforms can offer you and where. It does not determine how you should treat the activity.
If you are trading because you think you can price events better than the market — and you are tracking results, sizing positions deliberately, and passing on trades where your number matches the price — you are doing something closer to trading. Our expected value guide covers the arithmetic that separates the two.
If you are trading because a game is on and it makes the game more interesting, that is gambling with a better fee structure. Which is fine, as long as you know that is what it is, and size accordingly.
The regulatory label will be settled by courts and Congress. The label that matters to your account is the one you apply to your own behaviour.
Frequently Asked Questions
Are prediction markets legally gambling in the US?
Not under federal law as it currently stands. Kalshi and Novig hold CFTC designated contract market licences, which classify event contracts as regulated derivatives and pre-empt state gaming law. Several states dispute this, and the courts have split.
Which states say prediction markets are gambling?
Nevada, Massachusetts, Maryland and Ohio have ruled against the federal framing, and Arizona filed criminal charges against Kalshi in March 2026. Tennessee ruled in Kalshi's favour. Eleven states introduced legislation during 2026.
What is the Prediction Markets Are Gambling Act?
The Schiff-Curtis bill introduced in March 2026, which would classify event contracts as gambling by statute rather than leaving it to regulators and courts. It is not law.
How are prediction markets different from sports betting?
There is no house taking the other side, the margin is far lower — under 1% against 4–5% — and you can sell a position before the event resolves. Our full comparison is in prediction markets vs sports betting.
Do I pay gambling tax on prediction market profits?
It depends which treatment you use, and the IRS has not said. Most filers report as ordinary income; the gambling treatment is unfavourable because losses are capped. See our tax guide.
Can prediction markets ban me for winning?
No, and this is a genuine structural difference. A sportsbook profits when you lose and can limit or close winning accounts. An exchange matches you against other users and makes the same money either way.
