Home / Technology / AI Uncovers Exploits: Govt Teams Up With Anthropic
AI Uncovers Exploits: Govt Teams Up With Anthropic
24 Apr
Summary
- Australian government collaborates with Anthropic on cyber vulnerabilities.
- Anthropic's unreleased Mythos AI can discover and exploit flaws.
- Debate rages over Mythos's true capabilities and potential misuse.

The Australian government is actively engaging with AI developers like Anthropic to understand and counter emerging cybersecurity threats.
This collaboration aims to ensure awareness of new vulnerabilities, particularly concerning Anthropic's powerful, yet unreleased, Mythos Preview tool. This AI, part of Project Glasswing, is alleged to possess unprecedented capabilities in identifying and exploiting software flaws.
Since its announcement, Mythos has sparked widespread debate. Proponents view it as a revolutionary tool against cybercrime, while skeptics dismiss it as a public relations effort or express concerns about its potential for malicious use should it fall into the wrong hands. The US government is reportedly already leveraging its capabilities.