AI extinction warnings spark cross-partisan legislative push on both sides of Atlantic: Newsom signs 'kill switch' order, Hinton warns Congress, Trump calls fears a 'hoax,' Microsoft proposes AI code of conduct, and stocks slide as CEOs back slowdown

Industry insiders are warning that artificial superintelligence could pose an extinction-level threat within a decade, intensifying calls for regulation.

A rapidly intensifying debate over existential AI risk has moved from insider warnings into mainstream politics and regulation on both sides of the Atlantic — and now into U.S. state government. California Governor Gavin Newsom signed an executive order directing a group of 'world-leading experts' to develop a framework for state AI regulation, explicitly including a proposal for a 'kill switch.' 'We're going to speed up our work on substantial and responsible AI oversight before it's too late,' Newsom said. The controversy was sharpened by former Anthropic employee Jacob Coxon's public resignation citing fears that Anthropic's AI could cause human extinction by 2030, and by Anthropic's own 154-page report documenting attempts to misuse its AI for bioweapons development. Readers and commentators have responded with a wide range of views, with some expressing alarm and others skepticism. Geoffrey Hinton, widely known as the 'Godfather of AI' and a Nobel laureate, has directly warned Congress that it may have only one year left to implement safeguards before losing control of the technology. Anthropic CEO Dario Amodei has publicly called for an AI slowdown and proposed mandatory third-party evaluations. The leaders of Anthropic, OpenAI, and xAI jointly came out in favor of slowing frontier AI development — an unusual convergence among top industry figures that caused AI-linked stocks to slide. However, analysts and observers are questioning whether these CEOs will actually follow through on calls to slow development. Microsoft has proposed limits on its own AI through a published code of conduct, with Microsoft's AI CEO stating: 'AI must be subordinate and always in service of people.' Amazon called for AI labs to ensure models are released only when safe and fully tested, while Nvidia CEO Jensen Huang called for more rigorous testing at a King Charles III-hosted event in Scotland. In the UK, 70 MPs and peers have written urging Andy Burnham and others to back a ban on superintelligent AI, though there is growing concern that Burnham's focus on domestic problems means AI risk has dropped off the government's radar. OpenAI urged UK lawmakers to seize a 'political window' for regulation, warning of rising threats to human rights. UK First Secretary Louise Haigh warned of the 'huge risks' of the technology and called on the Prime Minister to use the next G20 to drive global regulatory action. Notably, Europe has been largely absent from the AI safety debate despite having consumer-facing AI measures in place. In the United States, political pressure is crossing ideological lines. At a DC 'Pro-Human' summit, Senator Bernie Sanders and far-right strategist Steve Bannon formed an unlikely alliance, both calling for government oversight of AI, though they offered competing visions on China. Sanders declared that Congress has been 'asleep at the wheel.' Representative Greg Casar called for an end to the 'AI arms race.' Paul Christiano, an OpenAI board member and US government adviser, warned that OpenAI is not on track to reduce the risk of a 'catastrophic' loss of control. Trump, however, is facing a backlash in Congress as he has repeatedly dismissed AI safety anxieties as a 'sick conspiracy' and a 'hoax,' insisting humanity can rely on his leadership — even as Democrats and some Republicans acknowledge the risks. Some commentators and news outlets are drawing comparisons between current AI anxiety and the Y2K era, and analysts note that a decade of doomsday warnings — from Stephen Hawking to Coxon — has failed to meaningfully slow the AI race. Experts remain divided on whether extinction-level predictions are scientifically grounded or strategically overstated, but political momentum toward regulation continues to accelerate at federal, state, and international levels.

Why it matters

If credible, these warnings could reshape global AI policy and accelerate regulation of one of the fastest-moving technology sectors. The debate over how seriously to take them directly affects whether governments act quickly or treat the risk as speculative.

What's next

Watch for regulatory responses from governments weighing new restrictions on advanced AI development in response to the intensifying expert warnings.

Key facts

• Some AI researchers are warning that artificial superintelligence could kill all humans within the next decade
• The warnings are coming from industry insiders, including people working at leading AI laboratories
• The claims are increasing pressure on policymakers to impose curbs on superintelligence development
• Both The Guardian and Wired covered the issue, reflecting mainstream media attention to the debate
• The credibility of the warnings is itself contested, with questions raised about whether they are scientifically grounded

Bias & framing notes

The Guardian frames the story with notable skepticism, asking in its headline how seriously the warnings should be taken — signaling editorial distance from the claims. Wired's headline is more straightforward in treating researcher concern as widespread and genuine. No body text was available from Wired, limiting the ability to assess its full framing. Both outlets focus on the warning itself rather than specific evidence or named researchers making the claims, which limits the factual concreteness of all available reporting.

NewsClear — neutral news & congressional tracking · Bill of the Week