Khanna bill would ban recursive AI, create FDA-style agency

Representative Ro Khanna will introduce the Human Control Over AI Act, a bill that would bar AI systems that "recursively self-improve or autonomously modify their own core objectives, containment or shutdown controls" until a new federal agency licenses them, CNBC reported. The proposal would stand up an AI safety agency modeled after the FDA to oversee frontier models from OpenAI, Anthropic, Google DeepMind and xAI, with authority over sandbox testing, air gaps, kill switches, and controls meant to keep models from escaping lab settings onto the internet.

Khanna is framing the stakes in existential terms. "There's actually a civilizational extinction risk," he told CNBC, adding that "there's a safety risk of loss of control, and then there's a misuse risk, and we need to take both seriously." In separate remarks he likened an AI arms race to nuclear conflict, saying "winning an AI war is like winning a nuclear war," an outcome he called "catastrophic for civilization."

The bill would require frontier AI companies to carry extensive liability insurance before releasing models, and would impose criminal penalties on employees who disable safeguards, kill switches, logging or containment systems, or knowingly deploy an unauthorized system. It also creates a "crimes against humanity" category for the deployment of AI models that result in the destruction of civilian populations.

On the international front, it calls on the administration to pursue enforceable agreements and targeted export controls to deter China and other adversaries from developing dangerous AI systems. Khanna cast the effort as an attempt to elevate voices he says have been sidelined, arguing the AI safety community has been "unfairly dismissed as science fiction and haven't been taken seriously enough."