Canadian Prime Minister Mark Carney says a U.S. export ban that prevents foreign access to Anthropic’s newest AI models illustrates the risk of depending on a small number of large, external AI providers. Carney made the comments during a trip to Ireland, framing the situation as a broader “model risk” problem—where systems become vulnerable if a key model provider is suddenly restricted or unavailable.
Multiple outlets report that Anthropic takes its latest models, Fable 5 and Mythos 5, offline to comply with the restrictions, and that the United States blocks all foreign access to them. Carney argues that this kind of abrupt cutoff shows why governments and businesses should not rely exclusively on a limited set of powerful models from major suppliers. One report also likens the episode to systemic risk themes associated with past financial disruption, using it as an analogy for how concentrated dependency can spread impact when rules or access change.
The coverage agrees that the U.S. action is the immediate trigger, while Carney’s main message focuses on reducing reliance on a handful of dominant AI tools.
