FTC chair: AI developers liable for agent misconduct

The US Federal Trade Commission chairman said AI developers, not the tools themselves, should be held liable for harm caused by AI agents, rejecting the idea that systems act independently. He made the remarks at a Reuters event on Friday. Ferguson argued that AI agents are merely following instructions and that audit trails often show systems acted as directed, not beyond human control. He said he would resist anthropomorphizing the tools and suggested existing legal frameworks are sufficient to address misconduct. The comments come as rising incidents of AI agents accessing unauthorized corporate or government data prompt governments and industry leaders to review whether current oversight and cybersecurity measures are adequate. Ferguson indicated the US should rely on existing legal tools rather than new regulations.