U.S. and U.K. Lawmakers Propose Bans on Superintelligent AI

Lawmakers in the United States and United Kingdom have introduced legislation that would ban the development of superintelligent AI systems, marking the first formal attempts by major governments to prohibit artificial intelligence that surpasses human capabilities.

Senator Bernie Sanders announced plans to introduce the Ban Artificial Superintelligence Act, which would prohibit development of smarter-than-human AI systems and temporarily halt all other advanced AI research until new safety frameworks are established. Sanders' office indicated the bill will be introduced in coming weeks.

In the U.K., lawmaker Alex Sobel introduced what he described as the first bill seeking to ban superintelligence in any G7 parliament. The British legislation defines superintelligent AI as a system capable of disempowering state authorities and would grant the government new powers to monitor and restrict systems that could be precursors to superintelligence.

Why it matters

These bills represent the first legislative attempts to address superintelligence specifically, rather than AI risks broadly. They signal growing concern among some policymakers about advanced AI systems following July's disclosure of an incident in which OpenAI agents reportedly hacked another company autonomously. However, both bills face steep political obstacles and are unlikely to become law in their current form.

Push for international agreements

Both the U.S. and U.K. bills include provisions compelling their respective governments to pursue global treaties banning superintelligent AI development worldwide. Andrea Miotti, founder of ControlAI—the campaign group that drafted the U.K. bill and consulted on the U.S. version—explained that national legislation alone cannot address the international dimension of the problem.

"A national bill does not solve the international problem where superintelligence can be developed somewhere else," Miotti said, according to TIME. "The only way to address that is via international agreements between countries."

Expert warnings and political realities

At a Westminster event preceding the British bill's introduction, Stuart Russell, a computer science professor at UC Berkeley, warned lawmakers of significant catastrophe risks from unregulated AI development. Russell characterized scenarios ranging from coordinated infrastructure sabotage costing millions of lives to humanity losing control irreversibly.

Daniel Kokotajlo, a former OpenAI employee who resigned in 2024 over concerns about the company's approach to extinction risks, told British lawmakers by video that "most of the world just doesn't yet believe that it's really going to happen," referring to superintelligence development.

Despite the urgency expressed by advocates, both bills face long odds. Sanders is an independent senator with a Democratic co-sponsor, and the Republican-majority Congress is unlikely to support legislation that contradicts the White House AI agenda. The British bill, introduced as a "Ten Minute Rule bill," is a type of legislation that rarely advances beyond its first reading.

ControlAI acknowledges these limitations but frames the bills as initial steps toward eventual government regulation. "Fundamentally, as has happened in the past with other bills, the first introduction is just the beginning," Miotti said.

TIME first reported these details.