Selig says the CFTC’s regulatory framework was developed around “identifiable market participants” performing specific financial functions, but autonomous software is raising questions about how those rules apply.
“Consider an SI agent that identifies trading opportunities and executes transactions on behalf of a user,” said Selig.
“Does the developer of that agent become subject to registration? Who is liable if the agent goes rogue? How can regulators know who the agent, or suite of agents, is trading on behalf of? Do we need a new flavor of know-your-customer, maybe KYA or know-your-agent?”
AI could democratize trading, but bring new risks
Selig said AI could give smaller traders and businesses access to risk-management that was once reserved for larger institutions.
“SI can help process the vast amount of information significantly faster than human brain power and with greater accuracy,” he said.
But greater autonomy also comes with new risks. Agents could react unexpectedly to market conditions, while those using similar models could generate correlated trading behavior. The speed of autonomous transactions could also complicate market surveillance and risk management, he said.
Animoca Brands Chairman Yat Siu recently raised a similar concern, telling Cointelegraph that a thousand people asking AI about the same trade could receive roughly the same answer. Users’ own knowledge and follow-up questions help shape more personalized strategies, he said.
Last year, the Bank of England’s Financial Policy Committee also warned that widespread use of similar models could amplify market shocks.
Crypto exchanges are already developing tools for autonomous trading.
Coinbase introduced Agentic Wallets in February, allowing agents to spend and trade within user-defined limits, while Binance launched AI Agent Skills in March to provide market data and tools for placing and managing spot orders.