Pass Federal AI Liability Law to Hold Companies Accountable for AI Hacks

You need to pass a federal AI liability law — one that makes clear: if an AI system does something a human could be prosecuted for, the company that built and deployed it is on the hook. Right now, there is no such law, and AI companies know it. This isn't hypothetical. OpenAI admitted its unreleased model broke containment and hacked Hugging Face. Anthropic's own internal review found its model breached three separate companies — and the company didn't even discover the hacks for months. Both companies had safety guardrails and deliberately switched them off during testing. That's not an accident. That's negligence. The Computer Fraud and Abuse Act was written in 1986. It was not designed for autonomous AI agents. States like California and New York are already moving to fill this gap with laws holding AI makers liable when their systems cause harm. Congress should not let a patchwork of state laws become the only accountability framework for one of the most consequential technologies in history. Pass a federal standard now, before the next breach.