Australia Gets AI Model With Hacking Power

Australian organizations are gaining access to Anthropic's Claude Mythos Preview, an advanced artificial intelligence model previously deemed too dangerous for public release. This initiative is part of a global, staggered rollout known as Project Glasswing, designed to strengthen critical infrastructure against cyber threats. The AI model has proven exceptionally adept at discovering software flaws.

Anthropic has deliberately withheld Mythos from general release due to its potent cybersecurity capabilities, which carry dual-use potential and could facilitate attacks on sensitive sectors like financial systems and infrastructure. The expansion includes approximately 150 new organizations across more than 15 countries, notably within the Five Eyes alliance, NATO, and the EU's cybersecurity agency.