Anthropic Expands Project Glasswing, Increasing Access to Claude Mythos AI Model
- Anthropic is adding approximately 150 organizations to Project Glasswing for access to the Claude Mythos AI model.
- The expansion includes sectors like power, water, healthcare, communications, and hardware.
- Organizations using the model have identified over 10,000 high or critical-severity software vulnerabilities.
- Anthropic has filed confidentially for an IPO with the U.S. Securities and Exchange Commission.
- The U.K. AI Security Institute reported that Mythos completed a simulated corporate network attack autonomously in testing.
Anthropic’s expansion of Project Glasswing aims to enhance cybersecurity by allowing more organizations to test and improve the Claude Mythos AI model before its broader release. This initiative involves critical infrastructure sectors where potential cyberattacks could impact millions globally.
By identifying over 10,000 vulnerabilities, partners demonstrate the model’s effectiveness in bolstering security measures ahead of its public launch. (Source)
