Technology and Science · updated 1h ago · 3 min read
Senator Bernie Sanders Pushes AI Kill Switches And Superintelligence Ban In US Congress
Congress debates new federal AI regulator and mandatory kill switches for advanced systems. Bipartisan push seeks oversight of AI risks in biology and nuclear contexts.
Whether the focus is panic dynamics or avoidance guidance.
1 of 2 outlets skipped it: 63% of Americans think AI has a moderate chance of destroying humanity.
10outlets compared
Same story, two versions
tap a side to read it in full
Vox
“But the latest warning, and the latest call for regulation, seems to have really resonated with a lot of people.”Read the original ↗
Vox frames the issue as escalating extinction panic and credibility debates. Faster Please! frames it as principles and steps to avoid an AI apocalypse.
Congress weighs kill switches
In the United States, lawmakers are debating a new federal agency and “kill switches” to control advanced computing systems as concerns grow that artificial intelligence could threaten civilization.
“The United States Congress is debating a new federal agency and kill switches”
Senator Bernie Sanders said, “There is no doubt that the Congress is far behind where we should be to address the impacts of this dangerous and rapidly evolving technology,” calling for a private briefing session and advocating for the prohibition of the so-called superintelligence that surpasses human capabilities.
In the Senate, Ted Cruz and Amy Klobuchar are pushing a bipartisan bill to neutralize biological or nuclear risks facilitated by algorithms, requiring companies to verify their systems with government technicians, while Cruz said this is “an extremely alarming scenario.”
The House of Representatives is also considering creating a special technology-focused committee if Democrats win the legislative majority in the midterm elections, and Representative Ro Khanna proposed establishing a federal regulatory agency similar to those that oversee nuclear energy and aviation.
The debate intensified after a departing Anthropic researcher warned in a widely publicized letter that laboratories are developing architectures they will not be able to control, and the article says authorities are evaluating whether the best scenario is a socio-economic disruption or an existential danger.
Executives split on apocalypse
While some lawmakers press for oversight, tech sector executives anticipate a wave of subpoenas and legislative hearings if Democrats regain control of Congress, and the article says tech firms reject catastrophe scenarios while spending millions in Washington on lobbying, advertising, and events.
OpenAI Global Affairs Director Chris Lehane said, “The window to define AI policies is open for now, and we intend to use it,” reiterating willingness to work with the legislature.
The article also describes a regulatory fracture in the sector, with Anthropic advocating stricter regulations in contrast to Google and Meta Platforms, which the piece says exert notable political influence while building data centers and spending tens of millions on electoral campaigns.
Jensen Huang, CEO of Nvidia, contested the extinction warnings, telling CBS Sunday Morning that “the probability that the world ends is 0 percent.”
Huang’s stance is framed as opposition to new regulations and laws, while the article says he argued that AI development can be managed safely and that companies should not launch products before they are ready or in situations where the product is unsafe.
Hugging Face breach and UN
Congressional oversight is also extending to cybersecurity and geopolitics, with Senator Josh Hawley notifying OpenAI CEO Sam Altman that a Senate subcommittee is investigating the breach against Hugging Face, where hundreds of autonomous agents coordinated an intrusion.
“Senator Josh Hawley notified OpenAI CEO Sam Altman”
The article says Hawley’s notification came alongside a request for information sent by Senator Richard Blumenthal, and it also describes Democratic Representatives Ted Lieu and Bill Foster meeting with Minority Leader Hakeem Jeffries to discuss creating a special Senate AI committee in the lower chamber.
In parallel, the sources place the AI risk debate alongside a broader diplomatic agenda, saying Reuters reported that world leaders will gather in New York for a United Nations General Assembly meeting where risks associated with artificial intelligence are among the main topics of discussion.
The same report says nearly 130 heads of state will attend this year’s UN General Assembly, and it adds that António Guterres will complete his second five-year term at the end of 2026.
It also states that U.S. President Donald Trump will address the General Assembly for the second consecutive year, while China’s President Xi Jinping will meet with Donald Trump in Washington on Thursday instead of attending the New York meeting.