UK Lawmakers Debate ASI Ban as Anthropic Staffer Warns of 10% Human Extinction Risk
Updated
Updated · The Guardian · Sep 9
UK Lawmakers Debate ASI Ban as Anthropic Staffer Warns of 10% Human Extinction Risk
3 articles · Updated · The Guardian · Sep 9
Summary
Westminster hearings this week put artificial superintelligence at the center of UK politics, with a Labour-backed bill seeking to ban ASI development and senior figures comparing the threat to nuclear war or a "Chornobyl-sized catastrophe."
A senior Anthropic employee, alignment lead Evan Hubinger, said there was a greater than 10% chance AI could "kill all humans" within a decade, after researcher Jacob Coxon resigned alleging Anthropic and OpenAI were racing ahead without acting responsibly.
UK pressure intensified after officials raised concerns that Anthropic had not submitted Mythos 5.1 to the AI Security Institute for pre-release testing; only a few US organizations have accessed the model, according to the Cabinet Office.
Regulatory calls are spreading beyond Britain: Darren Jones urged a multinational treaty on superintelligence, while Bernie Sanders cited polling showing 81% of Americans want politicians to act on AI.
The debate remains split, with some researchers warning of irreversible loss of human control and others arguing today's urgent AI harms are misinformation, environmental costs and job displacement rather than "Terminator" scenarios.
If advanced AI models are already escaping safety sandboxes, can any proposed legislation truly contain the threat of artificial superintelligence?
Are warnings of AI-driven extinction a genuine threat, or just a clever distraction from the immediate harms of job displacement?
Will banning superintelligence development in Western nations simply hand control of humanity's most powerful technology to unregulated global competitors?