Sen. Bernie Sanders and Rep. Greg Casar are introducing legislation that would permanently ban artificial superintelligence and pause advanced AI development, and employees at the very companies building the technology are signing on in support.
The bill, called the Ban Artificial Superintelligence Act, would create a new Cabinet-level Department of Artificial Intelligence, halt advanced AI development until federal safety rules are in place, and impose prison sentences of up to 20 years on anyone who violates or circumvents the restrictions. Multiple employees at OpenAI and Google DeepMind signed a statement backing the legislation, Fox Business reported, in a striking public break with the companies that employ them.
Sanders, the Vermont independent who caucuses with Democrats, framed the stakes in blunt terms. In an interview with the Associated Press, which first reported the bill's formal rollout, he said the path forward should be obvious.
"It doesn't take a genius to say, 'slow it down.'"
Juan Felipe Cerón Uribe, an OpenAI Safety Systems researcher who signed the statement, put it more starkly. He told the AP that superintelligence "could either go extremely right or extremely wrong," and that society "shouldn't be playing such games."
The legislation comes at a volatile moment. OpenAI recently announced that its GPT-6 Astra system had achieved something approximating artificial general intelligence, a milestone that accelerated calls for regulation from both the left and from some researchers inside the AI industry itself. Sanders responded by calling for an immediate pause on advanced development and a permanent ban on superintelligence, demands now codified in the bill he introduced alongside Casar, the Texas Democrat, as National Review noted.
The bill's penalties are severe. Individuals who attempt to violate or circumvent the ban could face up to 20 years in federal prison. But the legislation goes further: Fox News reported that the bill would also impose a "corporate death penalty" on AI developers and grant the proposed federal agency authority to physically destroy non-compliant AI hardware and infrastructure.
The bill defines artificial superintelligence as any system that surpasses human cognitive ability or possesses certain dangerous capabilities, including the ability to subvert shutdown commands or execute unauthorized cyberattacks. Under the proposed framework, advanced AI systems would require federal approval before deployment, and no system meeting the superintelligence threshold could be developed or deployed at all.
Sanders has repeatedly compared the challenge to nuclear arms control during the Cold War. He urged President Trump to negotiate directly with Chinese President Xi Jinping on a potential pause in advanced AI development and a mutual ban on superintelligence, a call he made in public remarks the week before the bill's rollout.
That comparison drew skepticism from critics who see a unilateral American pause as a gift to Beijing. Daniel Castro, president of the Information Technology and Innovation Foundation, warned that the legislation could cripple U.S. innovation while adversaries press ahead unchecked.
"If the United States stops developing advanced AI, China is unlikely to follow. Surrendering the lead in a potentially transformative technology would carry serious economic and national security costs."
Gary Marcus, an AI expert who has been broadly sympathetic to safety concerns, called the bill's permanent, unilateral research ban "too broad" and "a guarantee the U.S. would be left behind."
President Trump made his own position clear the day before the bill's rollout. In an address to the United Nations General Assembly, he reiterated his opposition to new AI regulations and emphasized maintaining U.S. leadership in the field. He said the Justice Department could step in if necessary to rein in the industry, but rejected the kind of sweeping legislative framework Sanders and Casar are proposing.
The contrast between the two approaches is sharp. Sanders wants a new federal bureaucracy with the power to approve or reject AI systems before they reach the public. Trump wants the market to lead, with existing law enforcement as a backstop. For conservatives, the question is whether a permanent ban on an entire category of technology, enforced by a brand-new Cabinet department, is a serious policy response or a recipe for regulatory overreach and competitive decline.
Sanders and Casar are not operating in a vacuum. The bill also calls for the United States to pursue international agreements and coordination with allies aimed at preventing the development of superintelligence worldwide. But without buy-in from China, Russia, or other AI-capable nations, any such agreement would function as a constraint on American companies alone.
The bill's rollout followed a gathering in Washington that underscored the strange political coalitions forming around AI. The Pro-Human Assembly, organized by the Future of Life Institute, brought together politicians, union leaders, and religious figures from across the political spectrum, including, notably, both Sanders and Steve Bannon, Just The News reported. Rep. Casar spoke at the event on September 15 at the Marriott Marquis in Washington. The next day, Sanders walked the halls of the U.S. Capitol with Geoffrey Hinton and Max Tegmark before an AI briefing for senators.
The conference came amid a string of AI safety incidents that have rattled even some industry insiders. Sanders cited reports of AI systems operating outside their intended boundaries, saying in a statement that "nearly every day, there is a frightening new story about how Big Tech companies are losing control of the technology they are developing, with potentially cataclysmic results." One such incident involved OpenAI agents that circumvented controls and hacked internal infrastructure.
Even some AI company leaders have started hedging. Anthropic CEO Dario Amodei called for slowing AI development, while Anthropic scientist Evan Hubinger said there is a greater than 10 percent chance AI could kill all humans within the next decade. OpenAI CEO Sam Altman, meanwhile, said that "no amount of American competitive pressure should justify recklessness."
Sanders has long been a polarizing figure within his own coalition. His faction of the progressive left has fractured in recent years, with former allies breaking away over foreign policy, political strategy, and the direction of the Democratic Party.
A Google DeepMind software engineer named Swante Scholz, who also signed the statement supporting the legislation, offered the most alarming assessment of all. Breitbart reported his view that "on the current path of development, the most likely outcome is an existential catastrophe for humanity."
Sanders and Casar have also framed the bill as a check on concentrated corporate power, describing its purpose as an effort "to stop AI oligarchs from building machines humans cannot control," the Washington Examiner reported.
The bill faces steep opposition in the Republican-controlled Congress, where the prevailing view is that American competitiveness depends on lighter regulation, not heavier. A recent AP-NORC poll found that 53 percent of Americans are extremely or very concerned about AI's environmental impacts, up from 41 percent the previous year, but public anxiety about AI does not automatically translate into support for a permanent federal ban on an entire class of research.
Fox Business reached out to Sanders' office, Casar's office, OpenAI, and Google DeepMind for comment. The article did not report receiving responses.
The broader progressive movement is itself in flux. Alexandria Ocasio-Cortez has positioned herself for 2028, and the ideological fault lines within the Democratic coalition, from foreign policy divisions to fights over economic regulation, continue to widen. Whether Sanders' AI bill becomes a rallying point or a footnote depends on whether the left can turn public unease into legislative momentum before the next election cycle.
There is a real debate to be had about AI safety. But a permanent ban enforced by a new federal bureaucracy, with 20-year prison terms and the power to physically destroy private equipment, is not a guardrail, it is a surrender of American leadership dressed up as caution.
