Anthropic chief tells G7 leaders to ‘resist the temptation to splinter’ over AI

{"text":[[{"start":12.65,"text":"Anthropic chief executive Dario Amodei told G7 leaders they must “resist the temptation to splinter” over the rollout of advanced AI tools, days after the US blocked his company from exporting its latest model."}],[{"start":26.950000000000003,"text":"Amodei made the comments in front of US President Donald Trump and other leaders on Wednesday, according to people familiar with the discussions, and drew support from arch-rival Sam Altman of OpenAI."}],[{"start":38.5,"text":"The Anthropic boss said at the G7 summit in Évian-les-Bains, France, that he sympathised with the efforts of countries to prevent AI from getting into the wrong hands but urged democratic nations to come together. Altman emphasised that cyberdefence tools should be provided to all the countries in the room."}],[{"start":58.15,"text":"Demis Hassabis of Google DeepMind joined Amodei and Altman at the meeting and the three executives called for US-led collaboration on development of AI models, while underlining risks, including on bioterrorism and cyber security, if democratic alliances fractured."}],[{"start":76.5,"text":"The US last week blocked Anthropic from exporting its Mythos and Fable models on national security grounds after the Trump administration received reports that safety guardrails could be bypassed."}],[{"start":88.9,"text":"The decision raised fears in both Europe and Silicon Valley that the Trump administration was willing to prevent non-US customers, even in friendly countries, from accessing the latest models."}],[{"start":100.7,"text":"European customers are concerned their economies could be harmed if they are unable to use cutting-edge technology, while US tech companies fear the loss of business."}],[{"start":null,"text":"

"}],[{"start":110.75,"text":"French President Emmanuel Macron said the Anthropic dispute had “clarified the stakes” for the US and its allies in the G7. He warned that if the US “from one day to the next can turn off the switch”, it would damage the multitrillion-dollar US companies leading the AI arms race."}],[{"start":128,"text":"Macron said discussions with G7 leaders, including Trump, had been fruitful although not yet conclusive on how to deal with frontier models. He called for “stronger regulation of artificial intelligence” and warned against the risk of “non-cooperation among democracies”."}],[{"start":144,"text":"“What we are going to build over the coming months is, in a way, a platform for discussion and co-operation among a number of democracies in response to the risks posed by artificial intelligence . . . in order to jointly define common standards,” he said."}],[{"start":158.75,"text":"Indian Prime Minister Narendra Modi also expressed concern about the Trump administration’s recent move to block Anthropic’s models, according to people familiar with the discussion. He added that democratic nations must have access to AI models to protect critical infrastructure "}],[{"start":175.15,"text":"Hassabis and Altman focused on technical co-operation, suggesting a technical standards body and a US-led evaluation forum."}],[{"start":183.55,"text":"Arthur Mensch, chief executive of French AI company Mistral, said one of the big concerns discussed was how countries could make sure they had a supply chain for AI technology they could control. "}],[{"start":195.05,"text":"“When you have an intertwined supply chain, are you sure that your counterparts can’t cut you off? That was mentioned multiple times mainly by non-US participants,” he said."}],[{"start":205.4,"text":"Additional reporting by Henry Foy in Brussels"}],[{"start":217.15,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1781770214_2013.mp3"}