Prediction markets have grown 2,100% in a single year, now counting more than 10 million active users and roughly $44 billion in contract volume. The uncomfortable question underneath the growth curve: is this betting, or infrastructure?
A prediction market is not a casino — there's no house liquidity taking the other side of every trade. It isn't sports betting in the entertainment sense either; the outcomes drive real price discovery rather than pure spectacle. And it isn't gambling in the classical sense, since the underlying events are objective rather than chance-based.
At the same time, a prediction market genuinely is a financial market — traders are putting real money on real-world outcomes. It functions as a price-discovery mechanism, aggregating dispersed information more efficiently than any single expert could. And it's increasingly being used as infrastructure: for elections, weather, geopolitical events and business forecasting, not just entertainment.
That tension defines the regulatory debate. The "betting" argument holds that if people are putting money on outcomes, it needs the same protections as any betting product: manipulation controls, anti-money-laundering checks, consumer protection and fraud prevention. The "infrastructure" argument holds that this is price discovery, not entertainment, and should be regulated like financial markets rather than gambling. One complication both sides have to reckon with: where the outcome of a contract is resolved by an AI oracle rather than a neutral "dealer," the resolution mechanism itself becomes a new manipulation surface.
Three broad regulatory paths are on the table. Treat these platforms like casinos, with strict licensing, position limits, age verification and full KYC/AML. Treat them like financial markets, with different rules for centralized platforms than for decentralized ones, alongside market-integrity controls like manipulation detection and circuit breakers. Or do nothing, which is roughly the current state — some platforms operate with formal approval, some in regulatory gray zones, and some fully decentralized and effectively unstoppable.
The $44B market proves the demand is real. The question is which category regulators eventually decide it belongs to.
My own expectation: a regulatory tightening cycle over the next 12–18 months aimed squarely at the platforms operating in gray zones, followed by licensed, casino-adjacent frameworks emerging for centralized players — while fully decentralized platforms simply move further offshore. Prediction markets should be regulated. But as financial infrastructure, not as gambling.