What Minnesota's Prediction Market Law Actually Says, and Who It Doesn't Reach

When Minnesota's governor signed H.F. 2000 into law in May 2026, the reaction across the prediction industry was immediate and largely focused on its most dramatic feature: the law made it a felony to operate a prediction market within the state. Within hours, the federal Commodity Futures Trading Commission filed suit to block it. The legal fight that followed drew significant press attention and became the most visible confrontation between state gambling authority and federal financial regulation the prediction industry has ever produced.

Lost in that drama was a quieter question with significant practical implications. What exactly does the statute say, and who does it actually cover?

The answer matters a great deal, because the law's reach is defined not by what it targets in spirit, but by what it says in text.

Minnesota H.F. 2000 defines a prediction market as "any system, platform, or service that allows consumers to place a wager on the outcome of a future event." That definition is doing most of the legal work in the statute. Before the new law reaches any company or platform, a court or regulator must find that the company operates a system, platform, or service, that it allows consumers to use it, and crucially, that those consumers place a wager.

The word "wager" is not defined separately in the statute. In Minnesota law, as in virtually every American state's gambling statutes, a wager has historically required three elements: consideration, chance, and prize. Consideration is the legal term for what you put in. In every context where Minnesota courts have examined wagering, consideration means something of value that the participant risks. A participant who risks nothing, because they have nothing at stake, does not place a wager in the conventional legal sense. The consideration element is absent.

On SafeBets, users make no deposit and place no bet. They receive 100 unicoins at signup, at no charge, and those tokens cannot be purchased on the platform. A user who makes a wrong prediction loses one unicoin from their free starting balance. They cannot lose money they never possessed. Under any ordinary reading of "place a wager," a user who has no capital at risk and who received their tokens at no cost is not wagering.

This is not a novel argument engineered for this particular statute. It is the application of longstanding wager doctrine to a new product architecture. Courts examining promotional contests, skill competitions, and employer prediction markets have consistently found that the absence of user-supplied consideration removes the activity from the scope of gambling law. What is new is that SafeBets is the first prediction platform of significant scale to be designed this way from the outset.

The statute reaches what it was written to reach. SafeBets was not written to be within that reach.

The Minnesota statute was drafted with wager-based platforms in mind, because all significant prediction platforms at the time of drafting were wager-based. The legislature's concern was genuine and legitimate: platforms that take consumer money on uncertain outcomes expose users to financial loss and attract the same harms associated with other forms of gambling. The statute addresses those harms by targeting the mechanism that creates them.

SafeBets does not employ that mechanism. The harms the statute targets, consumer financial loss on uncertain outcomes, do not occur on the platform because no consumer financial stake is ever at risk.

It is worth being honest about the limits of this analysis. Statutory text is not the only thing that determines regulatory outcomes. Regulators can and do interpret definitions expansively, particularly when a novel product achieves functional outcomes similar to a targeted activity through a different mechanism. A creative regulator could argue that unicoins constitute consideration because they have value on the secondary market even if they were received for free. A court could disagree with the plain-language reading and find that the legislative intent of the statute reaches platforms it does not textually describe.

None of that is impossible. What I can say with confidence is that the most natural reading of the statute text, the reading a court applying ordinary principles of statutory construction would reach, does not cover a platform where consumers place no stake. And that natural reading is consistent with how Minnesota courts have historically analyzed wagering in analogous contexts.

The CFTC preemption fight will eventually resolve the question of whether federal event-contract authority displaces state gambling law for wager-based platforms. That fight is Polymarket's and Kalshi's to win or lose. SafeBets is watching it with interest rather than anxiety, because the outcome of that fight does not determine SafeBets' right to operate. The statute's text does, and on the text as written, SafeBets is not a prediction market as Minnesota has defined the term.

This is not a loophole. It is what regulatory compliance looks like when you design the product around the harm rather than around the definition.

This article is provided for informational and educational purposes and does not constitute legal advice.