British-Canadian computer scientist; CEO and co-founder of Cohere
The White House has restricted access to Anthropic’s Mythos 5 and Fable 5 for foreign nationals, citing national security. Anthropic has suspended access to these models, creating winners for open-weight providers like Mistral and DeepSeek and prompting renewed calls for AI sovereignty in Europe. The debate centers on control versus access as nations weigh strategic dependencies on US AI infrastructure.
The G7 has agreed to assess frontier AI risks and explore a trusted partners scheme to give non-U.S. access to advanced models while safeguarding security. Leaders warn that access can be blocked and stress the need for international governance and sovereign computing power.
Industry leaders have urged a slower pace in AI development following warnings from an Anthropic researcher about risks to humanity. Leaders are pushing for safety standards, regulatory coordination, and potential speed limits on model progress, while facing competition, profits, and political pushback. Public offerings by frontier labs like Anthropic and OpenAI are also in consideration.