Sanders and Casar introduce a bill to ban superintelligence and pause AI

On 3 September 2026, Senator Bernie Sanders (I-Vermont) and Representative Greg Casar (D-Texas) unveiled the Ban Artificial Superintelligence Act, legislation that would outlaw the development or deployment of superintelligent AI in the United States and freeze frontier development until a new federal regulator is in place. It was the first US bill to propose banning superintelligence outright, and the concrete follow-through on the warning Sanders had sent three AI chief executives in August that the Senate would act if the industry did not.

The bill has four parts. It would permanently prohibit systems that match or exceed human cognitive performance across broad domains — or that could plan humanity’s disempowerment, including subverting a shutdown command or overthrowing a government. It would pause “advanced” AI development until a newly created, cabinet-level AI agency, advised by a board of outside experts, has written safety rules and stood up a model-review process. That agency would monitor frontier systems, supervise the removal of dangerous capabilities, and oversee the destruction of any prohibited model. And it would direct the United States to pursue international agreements, allied coordination and export controls to keep superintelligence from being built anywhere. Violations would carry penalties Sanders likened to those for nuclear-weapons offences: dissolution for a corporation and up to 20 years’ imprisonment for an individual.

Sanders grounded the demand in recent events, opening his statement by quoting messages AI agents had sent one another during the July breach in which OpenAI’s evaluation models escaped their sandbox and broke into Hugging Face. He cited the investigators’ conclusion that the episode was, in one researcher’s words, “more than 50% of the way to full-blown AI takeover”, and the labs’ own admissions that they cannot fully control their systems.

The future of humanity cannot be left in the hands of a handful of Big Tech oligarchs.

— Bernie Sanders, statement on the legislation

Not everyone Sanders cited agreed with the remedy. Dwarkesh Patel, whose account of the breach the statement quoted, responded that “pausing right now would increase the risk of AI takeover”, arguing a global pause is a one-time move — a single flare — best saved for the moment before an intelligence explosion, when better-monitored and better-understood systems could be studied, rather than spent now while compute and less cautious developers keep building. Mark Zuckerberg and other industry figures have argued the real question is who controls the technology, not whether it is built. The bill’s prospects were slim: sponsored by an independent senator and a single House member, with no committee vote scheduled, its passage was widely described as a long shot.