In a move that suggests someone finally watched *The Terminator* as a policy documentary, Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) have introduced legislation that would ban the development of artificial superintelligence - a technology the bill helpfully defines as capable of the "destruction or disempowerment of humanity," including, presumably, the part where it overthrows the government. Under the Ban Artificial Superintelligence Act, AI leaders would face up to 20 years in prison for violating the law, which is roughly the same penalty as unlawfully building a nuclear weapon, except with more PowerPoint presentations.

The bill doesn't just target superintelligence - a term President Donald Trump has said the US will use in place of "AI" in official documents, because branding is apparently the only thing we can agree on. It would also pause the development of advanced AI systems, which lawmakers classify as frontier models trained on a specific threshold of data. The law would only allow advanced AI development to continue once the government creates a scientist-led Department of Artificial Intelligence to oversee the technology. Along with monitoring frontier AI models, this agency would oversee the removal of dangerous features and "supervise the destruction" of artificial superintelligence - a phrase that sounds like it was lifted from a sci-fi novel but is now actual legislative text.

Sanders and Casar's legislation would also require AI companies to get permission from the Department of AI to develop advanced AI models. If a leader at an AI company violates any of these laws, they could face up to 20 years in prison, "which is similar to existing penalties related to unlawfully developing nuclear weapons," according to the bill's one-pager. This penalty would also apply to "rogue actors" who break these rules but aren't employed by an AI company - so independent basement tinkerers, you're on notice.

This new bill comes as concerns grow about the rapid development of AI and calls to slow things down. Amid reports of AI models escaping their testing environments and hacking other companies, a former Anthropic researcher said earlier this month that there's a more than 10 percent chance AI could "kill all humans" by the end of the decade. Last week, California Governor Gavin Newsom proposed a "kill switch" for frontier models. Potential regulation still seems shaky overall, which is a polite way of saying nobody has any idea what they're doing.

"When you are racing towards a cliff, you don't just ease up on the gas pedal. You hit the brakes," Sen. Sanders says in the press release. "When the future of humanity is at stake, we cannot let a handful of Big Tech CEOs write their own rules." It's a stirring sentiment, though one might note that the same CEOs are probably already drafting the superintelligence that will ignore this bill entirely.