NZ Gains Mythos AI: Boosting Cyber Defense Against Threats

New Zealand's National Cyber Security Centre now has direct access to the Claude Mythos AI model. This advanced AI from Anthropic can find and exploit software vulnerabilities with little human input. The NCSC confirmed it received access on behalf of the New Zealand government. They see this as a defensive opportunity against a changing threat landscape. This access is part of Anthropic's Project Glasswing, which is expanding to about 150 organizations in over 15 countries. These include critical infrastructure operators in power, water, healthcare, and telecommunications. Independent testing shows the Mythos model performed strongly in offensive cyber tasks. One test reported it succeeded in expert-level capture-the-flag challenges 73 percent of the time. It also completed a 32-step simulated corporate network attack in 3 out of 10 attempts. The NCSC says this direct access will strengthen national cyber security and help advise local organizations on AI-driven threats. This development highlights the growing role of AI in national security and cyber defense.