CFTC Chair Says AI Trading Agents Could Break Rules Built for Humans

CFTC Chair Michael Selig told the agency’s first Frontier Forum in Washington on Wednesday that it is no longer far-fetched to picture a refrigerator buying groceries with money earned from staking Ether, or even launching its own memecoin. He argued that regulators need to adjust to a market where AI agents can hold assets and make trades without human input. Selig also repeated President Trump’s description of the technology as “super intelligence,” likening its impact to the printing press, the steam engine and the internet.

Questions Over Liability and Know-Your-Agent Rules

Selig said the CFTC’s framework was built around identifiable market participants, and autonomous software strains that model. He asked whether developers of trading agents should have to register, who is liable if an agent goes rogue, and how regulators can tell whom an agent is trading for. He even floated a new idea called know-your-agent, a twist on know-your-customer checks.

Benefits and Risks of AI Trading Tools

Selig believes AI could give smaller traders and businesses risk-management tools once limited to big institutions. Still, agents may react unpredictably, and models that think alike could produce correlated trades. Fast execution could also make surveillance and risk control harder. Animoca Brands Chairman Yat Siu raised a similar worry, noting that a thousand people asking AI the same question may get nearly the same answer. The Bank of England warned last year that widespread use of similar models could amplify market shocks.

Growing Demand and Exchange Tools for AI Agents

An eToro survey from September found 79% of US millennial investors and 77% of Gen Z investors already use or are open to using AI for investing. Coinbase introduced Agentic Wallets in February, letting agents spend and trade within user-set limits, while Binance launched AI Agent Skills in March for market data and spot order tools.