The US government can shut off access to AI at will. What does this mean for Australia?

Anthropic faced an “export control” directive from the US government to block access to two AI models for foreign nationals.
Anthropic, unable to verify users’ citizenship, shut down access for all users globally, including Australians, as a stopgap measure.
The directive highlights the lack of sovereign control over foreign-hosted AI, emphasizing Australia’s reliance on US-based AI services.
Australia is urged to develop its own sovereign AI strategy to avoid being at the mercy of other nations’ export controls.
Achieving AI sovereignty involves data, compute, AI model, and policy sovereignty, with sustainable practices essential for community support.
Success in developing sovereign AI can adopt open-source models like Switzerland’s Apertus, complying with relevant regulations.
The EU’s progress in AI policy and investment provides a model for Australia to build independent and sustainable AI capabilities.