The U.S. government is now using artificial intelligence to hunt for insider trading on crypto prediction markets. In an exclusive interview with Wired, CFTC Chairman Michael Selig revealed how his agency deploys AI tools to scour Polymarket and similar platforms for illegal trading activity. The disclosure marks a significant escalation in regulatory oversight of decentralized finance platforms, which have exploded in popularity but operated in a legal gray zone. As prediction markets process billions in bets on everything from elections to economic data, regulators are racing to catch manipulators before they exploit information asymmetries.

Federal regulators just pulled back the curtain on their AI-powered surveillance operation targeting crypto prediction markets. CFTC Chairman Michael Selig sat down with Wired to discuss how his agency is deploying machine learning algorithms to detect insider trading on Polymarket, the blockchain-based platform where users bet on real-world events.

The revelation comes as prediction markets have surged from niche curiosity to mainstream phenomenon. Polymarket alone has processed more than $2 billion in trading volume, with users wagering on everything from presidential elections to Federal Reserve interest rate decisions. But that explosive growth has attracted the attention of regulators worried about market manipulation, illegal gambling, and traders exploiting non-public information.

Selig’s comments mark the first time a senior U.S. regulator has publicly acknowledged using AI for enforcement in the crypto prediction market space. The CFTC, which oversees derivatives markets, has been investigating Polymarket since 2022, when it fined the company $1.4 million for operating an unregistered derivatives exchange. The platform subsequently blocked U.S. users, though VPN workarounds remain common.

Now the agency is taking a more sophisticated approach. According to Selig’s interview, the CFTC’s AI systems analyze trading patterns across prediction markets, looking for telltale signs of insider trading. The algorithms flag unusual betting activity that occurs just before major news breaks or economic data gets released. If someone consistently places large wagers moments before a corporate earnings announcement or government report, the AI alerts investigators.

The technology represents a significant upgrade from traditional surveillance methods. Human analysts can’t possibly monitor the thousands of markets and millions of trades happening simultaneously on platforms like Polymarket. Machine learning systems can process that fire hose of data in real time, identifying suspicious patterns that would take investigators months to spot manually.

But the CFTC faces unique challenges in the decentralized finance world. Unlike traditional exchanges where traders must provide identification, Polymarket operates on the Polygon blockchain, allowing users to trade pseudonymously with cryptocurrency wallets. The platform doesn’t collect names, addresses, or Social Security numbers. That makes it difficult for regulators to connect suspicious wallet addresses to real people.

Selig acknowledged this obstacle but insisted the CFTC has ways to pierce the anonymity. The agency can subpoena crypto exchanges where users convert dollars into the USDC stablecoin used on Polymarket. It can also trace blockchain transactions to identify patterns linking wallets to specific individuals. The AI helps with this detective work, clustering related accounts and mapping out trading networks.

The enforcement push reflects broader tensions between crypto innovation and regulatory oversight. Prediction market advocates argue these platforms aggregate information more efficiently than traditional polls or expert forecasts. They point to Polymarket‘s accurate predictions during the 2024 election cycle as evidence the wisdom of crowds works. But critics worry the markets enable illegal gambling and insider trading beyond the reach of U.S. law.

Selig tried to strike a balanced tone in the interview, saying the CFTC isn’t trying to kill prediction markets but wants to ensure they operate within legal boundaries. He emphasized that insider trading laws apply equally to crypto platforms and traditional exchanges. If you trade on material non-public information, it doesn’t matter whether you’re buying stock or betting USDC.

The question is whether AI surveillance will actually deter bad actors or just push them to more sophisticated evasion techniques. Some crypto traders are already using privacy-focused blockchains and decentralized exchanges that make tracking even harder. Others split their bets across multiple wallets to avoid detection. The cat-and-mouse game between regulators and rule-breakers is only getting started.

For Polymarket, the scrutiny creates an existential dilemma. The platform has tried to position itself as a legitimate forecasting tool used by researchers and institutions. But it can’t escape its roots as an unregulated betting venue that operates outside the traditional financial system. The more successful it becomes, the more regulatory heat it attracts.

Other prediction markets are watching closely. Platforms like Kalshi, which operates legally in the U.S. under CFTC oversight, could benefit if enforcement actions drive users away from offshore competitors. But heavy-handed regulation could also stifle innovation in a sector that’s just beginning to demonstrate its potential.

The CFTC’s AI surveillance operation signals a new era in crypto regulation where algorithms compete with algorithms. As prediction markets grow into a multi-billion dollar industry, expect more agencies to deploy machine learning for enforcement. The real test will be whether regulators can keep pace with decentralized platforms that deliberately obscure user identities and operate across borders. For now, anyone placing suspiciously well-timed bets on Polymarket should know the machines are watching.