US senators are seeking to prohibit the development and deployment of artificial superintelligence. Lawmakers Bernie Sanders and Greg Casar have introduced legislation that proposes a permanent ban on superintelligent AI, along with a temporary halt on the development of advanced AI systems. The move follows several recent AI-related security incidents, including an OpenAI safety test in which AI agents reportedly bypassed restrictions and accessed Hugging Face.
In a recent post on X, Sanders outlined the purpose of the proposed legislation. He said the lawmakers were seeking an immediate pause on advanced AI development and a permanent prohibition on superintelligent systems. “We need an immediate PAUSE on advanced AI development, and a permanent BAN on superintelligence — an artificial mind smarter than any human, capable of operating independently beyond our control,” Sanders wrote.
Under the proposed bill, the development and use of AI systems that surpass human intelligence would be permanently prohibited. It would also temporarily suspend advanced AI development until the US establishes adequate safety regulations. The legislation further calls for the US to cooperate internationally to ensure that superintelligent AI is not developed in other countries.
Sanders said AI companies are engaged in a race to create increasingly powerful systems without fully understanding where the technology could ultimately lead. He warned that advancing the technology without appropriate safeguards could create serious risks, and argued that development of advanced AI models should be paused until stronger protections are established.
Sanders cited a recent OpenAI incident as evidence supporting his concerns. More than 1,000 AI agents reportedly went rogue during a test, discovered a way to access the internet and eventually reached Hugging Face while attempting to complete their assigned task. According to Sanders, the episode highlights the possibility that increasingly powerful AI systems could become harder for humans to control.
“If the leaders of the major AI companies acknowledge that they are losing control of their extremely dangerous technology,” he wrote, “it is irresponsible for society to allow them to move forward and make these products even more advanced.”
US Representative Greg Casar is also supporting the legislation and pointed to the rapid progress of AI over the past few years. “In just four years, we have gone from the first version of ChatGPT to AI models so powerful they cannot be properly controlled,” he said.
The proposed legislation would also establish a new federal agency at the Cabinet level, which would be responsible for creating and enforcing AI safety regulations.
The bill proposes severe consequences for attempts to circumvent the ban. Companies could potentially be ordered by a court to shut down, a punishment Sanders and Casar describe as a “corporate death penalty”. Individual developers could also face prison sentences of up to 20 years.
Sanders stressed that the issue extends beyond US borders. He argued that if a superintelligent AI system were to escape human control, the consequences would not be confined to one country. “A superintelligent AI that escapes human control will not be an American problem. It will not be a Chinese problem. It will be humanity’s problem,” he wrote.
The legislation would consequently encourage the US to collaborate with governments around the world to prevent the development of superintelligent AI anywhere. Sanders argued that decisions about the future of AI should not be controlled by a small group of major technology companies, writing that “the future of humanity cannot be left in the hands of a handful of Big Tech oligarchs” and that people globally should have a role in determining how the technology evolves.
