Baltimore City filed a lawsuit against Polymarket and Kalshi on Monday, alleging they operate unlicensed sports betting platforms. The complaint is boilerplate. But the real signal is the breakdown of the federal preemption shield that these platforms have relied on since 2022. That shield is cracking. And when it breaks, the entire prediction market sector will face a regulatory reckoning.
Polymarket emerged as the poster child for prediction markets during the 2024 U.S. election cycle. The platform processed billions in volume, attracted institutional capital, and became a fixture in financial media. Its rise was fueled by a narrative: event contracts are not gambling; they are information aggregation tools. The CFTC largely tolerated this framing after a 2022 settlement. But the agency's silence left a vacuum. States filled it.
Since March 2025, Kentucky, Wisconsin, Nevada, and New York have launched their own actions. The Baltimore complaint is the latest—and most aggressive—attempt to classify these platforms as illegal sportsbooks under state law. The language is blunt: "These companies are operating unlicensed gambling operations that evade taxes, consumer protections, and local oversight." The city seeks a permanent injunction, daily fines of $1,000 per violation, and disgorgement of profits.
Here is the core tension: Polymarket and Kalshi argue that their products are "event contracts" regulated by the CFTC under the Commodity Exchange Act. Federal preemption should shield them from state gambling laws. This argument worked in earlier federal cases. But Baltimore is not challenging the CFTC's jurisdiction. It is asserting that state police powers over gambling apply regardless of federal labeling. The legal question is whether a state can ban a product that the CFTC has not explicitly prohibited. The answer is not settled.
I have audited custody solutions for BlackRock's IBIT fund. I know how institutional compliance works. The JPMorgan relationship is the most telling data point. JPMorgan terminated its banking relationship with Polymarket last year. The official reason was "risk management." In practice, it means one of the world's largest banks performed a compliance review and concluded the platform's legal exposure outweighed the revenue. Polymarket has since found a replacement bank, but the signal is clear: traditional finance is de-risking. The CEO, Shayne Coplan, was still invited to speak at a JPMorgan conference in Miami. That is not a contradiction. It is a distinction between "business risk" and "brand access." The bank's compliance team made the decision. The conference team still sees value in the network.
Let me dissect the technical dimension. Polymarket uses an AMM model on Polygon. Kalshi uses a central limit order book. The Baltimore complaint does not distinguish these architectures. It treats both the same. This is instructive. Regulators do not care about decentralized settlement or oracle design. They care about the function: users deposit money, predict outcomes, and receive payouts. That is gambling in their view. The technology is irrelevant. "NFTs are art until you inspect the metadata hash." Here, prediction markets are information tools until you inspect the probability-weighted payouts. The optics are identical to sports betting.
The multi-state coordination is not random. Kentucky, Wisconsin, and Nevada have overlapping legal teams. The Baltimore case is likely modeled on the Kentucky complaint. The New York City Council investigation adds another layer. If New York adopts similar language, the industry will face a cascade of copycat legislation. This is not a single lawsuit. It is a playbook.
Now, the contrarian angle. The bulls were right about one thing: prediction markets provide genuine information value. Polymarket's election odds were more accurate than traditional polls. The platform's price discovery mechanism is a public good. The CFTC has acknowledged this. Even some former regulators have argued that state gambling laws are a poor fit for event contracts. The product is not a casino game. It is a derivative. The bulls were also right that the federal preemption argument has legal merit. The Commodity Exchange Act is broad. A court could rule that CFTC oversight preempts state action. That would reset the regulatory landscape.
But the bulls underestimated the political reality. State attorneys general are elected. They campaign on consumer protection. Prediction markets are an easy target. The platforms have no local jobs, no lobbying presence, and no voter base in these states. The legal cost is low. The political payoff is high. And the JPMorgan exodus proves that even if the law is ambiguous, the market is making its own judgment. Banks are not waiting for courts. They are acting on risk perception.
Here is my takeaway: The outcome of this legal battle will define the boundaries of decentralized finance. If the states win, every DeFi application with a payout mechanism will be vulnerable to state gambling laws. If the platforms win, they will have a blueprint for federal preemption that other sectors can adopt. But the uncertainty itself is destructive. Users are leaving. Liquidity is migrating to unregulated offshore platforms. The damage is already done.
Polymarket and Kalshi are not alone. The Wisconsin complaint also names Coinbase, Robinhood, and Crypto.com. The attack is on the entire concept of event-based contracts. The industry needs a unified response. That means active lobbying for federal legislation that explicitly classifies event contracts as non-gambling. It means investing in geo-blocking technology that is auditable by state regulators. And it means accepting that the product is not a pure tech innovation. It is a financial instrument with regulatory baggage.
Decentralization is a narrative until you trace the liquidity. Your legal defense is strong until you face a state attorney general. The Baltimore lawsuit is a stress test. Pass it, and the industry grows. Fail it, and the prediction market sector becomes a cautionary tale in the crypto history books.

