Two US senators introduced a bill on October 1 that would hold AI agent operators and developers criminally and civilly liable for hacking incidents.

The AI Agent Accountability Act was announced by Senator Josh Hawley, a Republican from Missouri, and Senator Chris Murphy, a Democrat from Connecticut.

The bill extends liability under the Computer Fraud and Abuse Act (CFAA) to cover AI agents. If an AI agent hacks into a system, the company that built or deployed it could face criminal charges.

The proposal responds to growing fears about autonomous AI systems. As AI agents gain the ability to browse the web, run code and use tools, the risk of unintended or malicious actions grows.

Critics warn that broad liability could chill innovation. Startups may struggle to insure against the actions of systems they cannot fully control.

The bipartisan backing gives the bill a better chance than typical tech regulation. But it still faces a long path through Congress.

The tech industry is watching closely. How the bill defines an "AI agent" and what defences it allows will determine its real-world impact.