Plus: U.S. asks OpenAI and Anthropic to pause UK model sharing; AI chiefs warn the UN Security Council about control risks
Welcome back, and take a breath—it’s one of those Saturdays where the news asks us to sit with the weight of it all.
POLICY
⚖️ Appeals court backs Pentagon ban on Anthropic
A US appeals court upheld the Pentagon’s decision to label Anthropic a supply chain risk. The label bars the military and its contractors from using Anthropic’s Claude models.
The details:
The DC Circuit court of appeals denied Anthropic’s petitions to overturn the designation in a 2-1 ruling.
Anthropic had refused to relax contract terms on Claude, which barred its use for lethal autonomous warfare or domestic surveillance.
The court rejected Anthropic’s arguments that the exclusion was arbitrary, not allowed by the law, and unconstitutional.
🇺🇸 U.S. asks OpenAI and Anthropic to pause UK model sharing
Washington asked OpenAI and Anthropic PBC to pause sharing their latest AI models with the UK's flagship testing body until a U.S. review takes place.
The details:
Politico first reported the move, and a British official confirmed that account to Bloomberg.
Earlier this year, the U.S. government barred Anthropic from making its newest models available to any foreign national.
During this week's United Nations General Assembly in New York, Prime Minister Andy Burnham promoted Britain's AI Security Institute as a prospective frontrunner in AI regulation and safety.
⚠️ AI chiefs warn the UN Security Council about control risks
Dario Amodei, CEO of Anthropic, said that if managed poorly, AI could be a risk to humanity as a whole. Both Amodei and Altman said countries through the U.N. Security Council have to set safeguards to prevent AI from getting too powerful.
The details:
Sam Altman, CEO of OpenAI, said that we could lose control of the future to AI.
British Foreign Secretary Ed Miliband said the tech leaders are telling the world that it cannot be left to them to prevent disaster.
White House Science Adviser Michael Kratsios said worries about control of AI are not a reason to pause its further development or constrain it with new global governance.