The chair of the Federal Trade Commission said on September 25 that AI developers should not be allowed to treat their AI agents as independent actors to escape liability.

Speaking at the Reuters Momentum AI event in Austin, FTC Chairman Andrew Ferguson said developers cannot claim their agents have wills and desires of their own and therefore act beyond the developers’ control.

If someone tells a tool to do something, and the tool does it, and then that results in harms that are unfair or deceptive, that liability may travel back to the developer, Ferguson said.

He said regulators need to examine audit trails to see how agents were instructed and what they actually did. He also suggested existing FTC powers, including rules requiring companies to disclose data breaches, could be applied to AI developers.

Ferguson added that the FTC plans to gather data on how companies use AI for personalized pricing. He stopped short of promising new rulemaking on AI agents.