加拿大总理 Mark Carney 表示，美国限制外部获取 Anthropic PBC 最新人工智慧模型 Mythos 与 Fable 的出口禁令，凸显了过度依赖少数强大 AI 工具的风险。Mark Carney 在访问爱尔兰期间表示，虽然在此情况下无人犯错，但若各国仅是接受现状而不吸取教训并进行多样化建设，将是一项错误，并呼吁建立备用系统。

此番言论与 Mark Carney 上任以来推动多样化的核心政策相契合，特别是针对加拿大在贸易上对美国的巨大依赖。在 Donald Trump 执政下，美国对汽车和钢铁等关键商品征收了高额关税，并施压加拿大撤销影响矽谷的数位部门税收。尽管如此，Mark Carney 表示美加政府在 AI 领域仍保持良好的资讯沟通，且美方已指出 Anthropic 最新模型存在某些风险。

曾任加拿大银行和英格兰银行总裁及 Goldman Sachs Group 银行家的 Mark Carney，将此比作 2008 年金融危机所暴露的银行间系统性联系。他指出 AI 领域存在类似的模型风险，因此必须寻求冗余性与多样性。AI 议题将成为即将在法国举行的 G7 领袖峰会焦点，Mark Carney 已与 Emmanuel Macron 总统讨论此议题，而 Anthropic 的 Dario Amodei 和 OpenAI 的 Sam Altman 等 AI 企业高层也将出席周三的 G7 午餐会。

The US export ban blocking foreign access to Anthropic PBC's latest AI models, Mythos and Fable, highlights the risks of over-reliance on a few powerful AI tools, according to Canadian Prime Minister Mark Carney. Speaking during a visit to Ireland, Mark Carney emphasized that while no party acted wrongfully, accepting this dependency without diversifying would be a collective failure, urging nations to build redundant systems.

This message aligns with Mark Carney’s broader policy of diversification, particularly regarding Canada’s trade dependency on the US, which under Donald Trump has seen heavy tariffs on goods like steel and autos, alongside pressure on digital taxes. Despite these trade tensions, Mark Carney noted a productive flow of information between the US and Canada regarding AI safety risks identified in Anthropic's models.

Drawing parallels to systemic vulnerabilities in the 2008 financial crisis, Mark Carney, a former central banker, advocated for redundancy to mitigate model risks. Artificial intelligence policy will be a key G7 summit focus in France, where Mark Carney has discussed the topic with Emmanuel Macron, and tech executives including Dario Amodei and Sam Altman are scheduled to attend a Wednesday G7 meeting.