Polymarket Under Siege: The Data Trail of a Multi-State Regulatory Assault

The Baltimore City complaint, filed in June 2025, is a 47-page document that reads like a legislative audit. It cites 12 specific instances where Polymarket and Kalshi allowed residents to place wagers on college basketball games, NFL spreads, and even the outcome of local mayoral races. The city’s legal theory is simple: if it looks like a bet, settles like a bet, and is marketed like a bet, it is a bet. The two platforms call their products 'event contracts.' The city calls them unlicensed gambling. Data does not negotiate; it only reveals. And the data here reveals a pattern of regulatory escalation that began in Kentucky in March, accelerated through Wisconsin and Nevada, and now converges on Baltimore with a request for a permanent injunction and $1,000 per violation per day in penalties. To understand the magnitude of this shift, one must first map the industry’s trajectory. Polymarket launched in 2020 as a decentralized prediction market built on the Polygon sidechain, using UMA’s optimistic oracle for settlement. Its core innovation was an automated market maker (AMM) model that allowed users to trade binary outcomes without a central order book. During the 2024 U.S. presidential election, the platform processed over $3 billion in volume, becoming a mainstream source for real-time probability data. Wall Street funds, political campaigns, and news outlets began citing its odds. But the same infrastructure that made it efficient—global accessibility, pseudonymous wallets, and instant settlement—also made it a target. By early 2025, the CFTC had already fined Polymarket $1.4 million for failing to register as a designated contract market. The company settled, agreed to block U.S. users from certain categories, and continued operating. The state-level assault, however, was just beginning. The core of the current crisis is a legal doctrine known as federal preemption. Polymarket and Kalshi argue that because their event contracts fall under the Commodity Exchange Act and are subject to CFTC oversight, state gambling laws cannot apply. This argument succeeded in a previous case involving Kalshi’s congressional control contracts. But the Baltimore complaint, along with actions in Kentucky, Wisconsin, and Nevada, deliberately sidesteps the federal securities question. Instead, they focus on the state’s traditional police power to regulate gambling. The city’s petition states that the platforms ‘operate without the required license, thereby avoiding the taxes, audits, and player protection obligations that licensed sportsbooks must bear.’ This is a forensic claim: the platforms are not just breaking the law—they are undermining the regulatory compact that funds addiction treatment and consumer protection. Based on my audit experience with similar compliance frameworks, this is a strategically powerful narrative. It paints the platforms not as innovators, but as free riders on a system designed to protect vulnerable populations. Let me dissect the technical response available to Polymarket. Geo-blocking is the most immediate tool. The platform can implement IP-based restrictions, require KYC verification for U.S. users, and even blacklist wallet addresses known to originate from litigious states. However, the effectiveness of these measures is limited. During my 2021 blind box audit failure, I learned that static analysis and technical barriers are insufficient against determined users. VPNs, proxy chains, and even physical SIM cards can bypass IP blocks. More critically, the Polygon blockchain itself is immutable and permissionless. Once a contract is deployed, there is no way to prevent a user from interacting with it through a self-custodied wallet. The only real lever is the front-end interface and the fiat on-ramp. If Polymarket shuts off its web interface and its banking relationships, the platform becomes a ghost protocol—functional but inaccessible to the average user. JPMorgan’s decision to terminate the banking relationship, reported by the Financial Times, is the most concrete signal that the financial system is already de-risking. The bank did not cite a specific law; it simply concluded that the regulatory cost of serving Polymarket exceeded the revenue. Data does not negotiate; it only reveals. And the data here reveals a liquidity squeeze that is not about code but about capital. The contrarian angle is that the bulls had a point. Federal preemption has worked before, and the CFTC has shown a willingness to accommodate prediction markets as a legitimate price-discovery mechanism. The Kalshi precedent, in which a federal court allowed the platform to list congressional control contracts, suggests that the courts may not automatically side with state regulators. Furthermore, the information value of prediction markets is real. During the 2024 election, Polymarket’s odds consistently outperformed traditional polling in accuracy. This is a utility that regulators should want to preserve. But the very success of the platform has created a new vulnerability. By becoming a mainstream data source, it attracted the attention of state attorneys general who see it as a threat to their regulated gambling industries. The JPMorgan relationship is a case in point. The bank’s decision to terminate the account was not based on a legal order but on a risk assessment. Chief Compliance Officer, speaking at a Miami conference, hinted that the bank’s internal scoring system flagged Polymarket due to the ‘concentration of legal actions across multiple jurisdictions.’ This is a classic pattern: once the regulatory noise reaches a certain threshold, the banking system self-censors. The irony is that the bulls are correct about the underlying value, but the market is not pricing in the speed at which the banking infrastructure can withdraw. What does this mean for the prediction market ecosystem? The most likely outcome is a bifurcation. Platforms that can afford to litigate and lobby—like Kalshi, which has a CFTC license and institutional backing—will survive, but at the cost of becoming de facto regulated sportsbooks. Platforms that cannot, like smaller offshore competitors, will retreat to the shadow of unregulated, permissionless protocols. Polymarket sits in the middle. It has the brand and the volume to fight, but it also has the most to lose. The Baltimore lawsuit, if successful, will create a template that other cities can copy. The injunction request is specifically designed to be scalable: a single judgment can be applied to any future user in the jurisdiction. The $1,000 per day penalty is low enough to be constitutional but high enough to be painful over time. This is a legal strategy that relies on volume, not severity. Data does not negotiate; it only reveals. And the data here reveals a systematic, multi-jurisdictional effort to force prediction markets into the existing regulatory box. In my forensic analysis of the Terra-Luna collapse, I observed that the illusion of liquidity was maintained until the circular trading patterns were exposed. Here, the illusion of regulatory safety is maintained until the state-level actions become too numerous to ignore. The JPMorgan termination is the equivalent of the first bank run—it signals that the traditional financial system is no longer willing to provide the rails. Polymarket’s CEO, Shayne Coplan, can still speak at industry conferences, but the operational reality is that the company is now fighting a war on five fronts: legal, regulatory, banking, technical, and public perception. The outcome will depend on whether the federal courts uphold preemption, or whether the states succeed in forcing a compliance-first model. Either way, the prediction market industry will never be the same. The question is not whether regulation will come, but whether it will be a single federal framework or a patchwork of state prohibitions. The answer will determine whether prediction markets remain a global, permissionless innovation or become a regulated, domestic utility. The choice is not a technical one—it is a legal one. And the courts are now the only venue where the data will be heard.