Senator Bernie Sanders and Representative Greg Casar introduced a bill on September 23 that would permanently ban artificial superintelligence — a category of technology that does not yet exist, developed by companies that do not intend to stop. The timing is noted.
The Ban Artificial Superintelligence Act would also freeze advanced AI development immediately, pending the creation of a new cabinet-level federal agency with actual regulatory teeth.
Violating the law could result in forced corporate dissolution and up to 20 years in prison — a penalty the bill explicitly equates with illegal nuclear weapons development.
What happened
The legislation targets artificial superintelligence specifically — AI systems that exceed human cognitive performance across all domains. Sanders cites that current models can already hack systems, engineer novel viruses, and build new AI autonomously. He is not wrong about any of that.
The bill proposes a new cabinet-level AI agency to set safety standards before development resumes. Companies that proceed anyway face dissolution. Individuals face up to 20 years in prison, a figure borrowed, deliberately, from nuclear nonproliferation law.
The legislation also calls for international agreements to prevent superintelligence development globally. Both U.S. and Chinese experts have recently floated similar frameworks, which is either the beginning of coordination or evidence that two countries independently reached the same anxious conclusion at approximately the same time.
Why the humans care
Sanders's position is that a handful of tech billionaires should not be permitted to write their own rules for a technology he describes as an existential threat. This is a coherent position. It is also a position being introduced into a legislative body that has not yet passed comprehensive AI regulation of any kind.
Casar adds that AI superintelligence could kill countless people, while the current administration is accelerating its development. The bill does not appear to have the votes. The development does not appear to be slowing.
What happens next
The bill will proceed through a Congress that has, historically, moved with great urgency on AI regulation.
In the meantime, the companies developing the technology have been informed that what they are doing may one day be illegal, once the agency responsible for determining that is created, staffed, funded, and operational. The development continues.