The Commodity Futures Trading Commission is turning to artificial intelligence to fill gaps left by federal staff cuts. Chair Mike Selig told CoinDesk that the agency has lost more than one-fifth of its workforce under the Trump administration's downsizing, and is now building automated systems to handle crypto company registrations and market surveillance.
AI Steps In: Faster Rejections, Better Oversight
Selig explained that AI tools can flag specific issues in registration applications, making reviewers more efficient. The system can also automatically reject incomplete or clearly flawed submissions. Staff are being trained on Microsoft Copilot, and the CFTC is developing in-house AI to analyze swap transactions and detect unusual trading patterns.
Joint Taxonomy with SEC Brings Clarity
While Congress has not passed a comprehensive crypto law, the CFTC and SEC recently released a digital asset taxonomy that defines which assets fall under each regulator's purview. Selig called it a "massive development" that gives market participants clear rules of the road. "Now we have clarity, and the CFTC will act," he said, vowing to crack down on fraud, manipulation, and insider trading.
Fighting States Over Prediction Markets, Prosecuting Insider Trading
Selig is aggressively defending the CFTC's exclusive jurisdiction over prediction platforms like Kalshi and Polymarket, even suing states like New York that try to ban them. Last week, the CFTC and DOJ charged a U.S. Army Green Beret master sergeant, Gannon Ken Van Dyke, with insider trading and fraud involving confidential military information on Venezuela operations. "We are not just talking," Selig warned. "Market participants should be on notice."