Sanders and Casar propose federal ban on AI superintelligence

Sen. Bernie Sanders and Rep. Greg Casar have introduced legislation that would prohibit the development of AI systems that exceed human cognitive abilities across multiple domains or pose existential risks to humanity.

The proposal, announced Wednesday by Sanders and Casar, who chairs the Congressional Progressive Caucus, represents one of the most restrictive regulatory approaches yet proposed by federal lawmakers. The bill would immediately ban so-called "superintelligence" while pausing other advanced AI development until a new federal Department of Artificial Intelligence establishes comprehensive safety rules.

Why it matters

The legislation marks a significant escalation in congressional AI regulation efforts, moving beyond voluntary industry commitments to mandatory development pauses and outright technology bans. If enacted, it would fundamentally reshape how AI companies operate in the United States and could influence international approaches to AI governance at a time when both the U.S. and China are racing to dominate the technology.

What the bill would do

The legislation would create a new cabinet-level Department of Artificial Intelligence, supported by a scientific advisory board. This department would establish binding rules for AI development, requiring companies to submit pre-development plans and obtain approval before deploying systems publicly.

Beyond banning superintelligence outright, the bill would also prohibit AI systems displaying "superintelligence precursor characteristics." These include capabilities to enhance nuclear, chemical, or biological weapons design, deceive humans at scale, or modify their own core functions autonomously.

The measure would also direct the new department to work with the State Department on international agreements to ban superintelligent AI globally.

Framing the risks

During a Wednesday press call, Sanders argued that allowing AI development to proceed without clear guardrails is "fairly insane" given that industry leaders acknowledge uncertainty about the technology's trajectory.

"If the leaders acknowledge they are losing control it is irresponsible in my view to allow them to move forward and make these products even more advanced," Sanders said.

He cited potential catastrophic events on the scale of the Chernobyl disaster or a global pandemic, as well as threats to democratic institutions and financial systems.

Casar emphasized the bill would not halt all AI development, noting enthusiasm for potential medical breakthroughs. He rejected arguments that a domestic pause should wait for international coordination, saying voluntary industry commitments based on companies' own risk tolerance "is not how this works in any other industry."

Political context

The announcement came during Chinese President Xi Jinping's state visit to Washington, with multiple lawmakers calling for U.S.-China cooperation on AI safety standards.

Rep. Ro Khanna, ranking member of the Select Committee on the Chinese Communist Party, urged the two nations to establish "basic inspections and safety standards," citing deep public concern about AI risks.

President Trump, speaking at the U.N. General Assembly earlier in the week, framed AI development as a competitive race and opposed what he called a "globalist scheme to control" the technology. Trump also said he prefers the term "superintelligence" to "artificial intelligence" because "artificial" makes it sound fake.

Brendan Steinhauser, CEO of the Alliance for Secure AI, praised Sanders and Casar for "taking superintelligence seriously" but stopped short of endorsing the ban, calling instead for "a more measured approach."

These details were first reported by Roll Call.