Australia’s largest private tech groups are treating the Trump administration’s weekend move against Anthropic’s powerful models as a geopolitical shock, not just a product change.
A new US export control directive instantly blocks access to Anthropic’s Mythos 5 and Fable 5 for all foreign nationals, including those physically inside the United States.
Officials in Washington frame the suspension as a national security measure, and point to reports that Fable 5 can be jailbroken to assist the kind of sophisticated hacking that already forced Mythos into a tightly controlled access programme.
Industry figures in Australia now warn that the country has hit a chokepoint for its most advanced AI capabilities, similar to an oil shipping lane suddenly closing.
Senior voices in Canberra highlight that decisions about who can use Mythos 5 or Fable 5 are being made in Washington boardrooms and agencies, not in Australian institutions.
The restricted access regime for Mythos had already limited which companies could experiment with its cutting-edge functions, and the Fable 5 reports intensified those limits.
Local AI builders and enterprise users are scrambling to reassess projects that quietly depended on Anthropic’s frontier tools.
Political leaders argue the episode lays bare how little sovereign capacity Australia has built in frontier AI.
They say the country leant heavily on US platforms while underinvesting in its own models and infrastructure, leaving national security judgments effectively outsourced to foreign providers.
The Anthropic ban is a live demonstration of what happens when critical digital systems sit outside domestic control.
