Bernie Sanders Moves to Ban AI Superintelligence. His New Bill Would Create a Federal Department of AI.
Sen. Bernie Sanders and Rep. Greg Casar have introduced one of Washington’s most sweeping attempts to regulate artificial intelligence, proposing a permanent ban on artificial superintelligence, a temporary halt to advanced AI development, and the creation of a cabinet-level Department of Artificial Intelligence.
The Vermont Independent and Texas Democrat introduced the Ban Artificial Superintelligence Act, arguing that the rapid development of increasingly capable AI systems is moving faster than the government’s ability to oversee them. “When you are racing towards a cliff, you don’t just ease up on the gas pedal. You hit the brakes,” Sanders said when announcing the legislation.
The proposal would permanently prohibit the development or deployment of “AI that exceeds human cognitive performance and capabilities across most domains, or has sufficient capabilities to destroy or disempower humanity, including by overthrowing the federal government.”
But the legislation would also reach beyond hypothetical superintelligent systems. It would also impose an immediate pause on the development of certain advanced AI “until a new, federal AI regulatory body is up and running and has established clear rules and model review processes to ensure safe and secure development and deployment of AI.”
At the center of that system would be a new Department of Artificial Intelligence. The cabinet-level agency would “safeguard the public from the dangers of artificial intelligence, including by enforcing a prohibition on artificial superintelligence.”
Those could include systems capable of evading shutdown commands or conducting unauthorized cyberattacks. The department would also have authority to oversee the destruction of prohibited artificial superintelligence.
The legislation comes with unusually severe enforcement provisions. Individuals who attempt to circumvent its restrictions could face prison sentences of up to 20 years. Companies could face what Sanders and Casar describe as a “corporate death penalty.”
Sanders and Casar are also calling for the United States to pursue international agreements, allied coordination and measures including export controls designed to prevent artificial superintelligence from being developed abroad.
The proposal is receiving support from some people working inside the AI industry. Juan Felipe Cerón Uribe, a researcher in OpenAI‘s Safety Systems, said in a statement supporting the legislation that superintelligence could go “extremely right or extremely wrong.”
Swante Scholz, a Google DeepMind software engineer speaking in a personal capacity rather than for the company, also backed a ban, warning of potentially catastrophic consequences from the current trajectory of AI development.
The bill faces substantial political obstacles in the Republican-controlled Congress, where lawmakers have yet to reach agreement on less restrictive federal AI rules. It also arrives amid arguments that slowing American AI development could give China an advantage in the global technology race.
Sanders argues that the answer ultimately requires international cooperation. He compared the challenge to Cold War nuclear arms control, pointing to agreements negotiated between the United States and the Soviet Union as evidence that competing powers can cooperate when a technology poses potentially global consequences.
“Ronald Reagan, arch-conservative, was smart enough to understand that he had to sit down with Gorbachev, a communist, and work out a nuclear treaty that protected humanity,” Sanders told the Associated Press. “I think we can do that again now.”
The legislation also marks an expansion of Sanders’ recent focus on AI. Earlier this month, Sanders and Casar announced plans for the superintelligence ban, while Sanders has separately pushed proposals addressing AI infrastructure and the economic effects of automation.
