CANADA PERSPECTIVE

Ottawa reads Anthropic's request with a strong local angle: the day before the announcement, University of Toronto researchers published work showing how AI tools can create a new type of computer "worm" able to adapt its hacking strategy as it spreads from device to device and takes over a vast network. Lead researcher Nicolas Papernot tells the National Post: "I think it's really important that people understand that it's not just the biggest, most powerful language models that pose the security concerns." For the Canadian press, this timing validates Anthropic's alert — but also reveals its partiality: Anthropic says "pause," but it is the Canadian research community that concretely documents the cyber risk. The National Post repeats Anthropic's key argument: "the evidence suggests that the human role is narrowing at each step in the AI development process." And keeps Anthropic's central metaphor: the situation is harder to control than nuclear non-proliferation treaties "because AI training is far easier to hide than a missile silo, and the temptation to quietly keep going would be enormous." Canada frames the stakes as a matter of technological sovereignty: a country that hosts one of the world's best AI schools (MILA Montréal, U of T) cannot be a spectator of a debate whose conclusions will be drawn in San Francisco.