Prime Minister Mark Carney says a U.S. export ban that bars foreign access to Anthropic PBC’s newest artificial intelligence models highlights the risk of depending on a small number of large AI providers. Multiple outlets report that the U.S. restriction effectively forces Anthropic to take its latest models offline to comply with the policy. Carney frames the episode as an example of “model risk,” arguing that governments and businesses should avoid building systems around a limited set of dominant tools from a few companies and countries.
Several reports add that Anthropic’s pulled models are named Fable 5 and Mythos 5, and that the U.S. action is tied to national-security-related export controls. Carney urges greater diversification of AI partnerships and technology sources to reduce systemic vulnerability if access is suddenly cut off. One outlet also includes Carney’s comparison of the situation to broader systemic risk, while others focus on his call for redundancy and wider sourcing. Overall, the coverage centers on Carney’s message that restrictions on leading AI model providers can have immediate downstream effects for users outside the U.S.
