Anthropic introduced Project Glasswing, an initiative focused on using AI to detect and prevent AI-driven cyberattacks. The project brings together security experts and researchers to evaluate risks from advanced AI systems and develop safeguards. Anthropic aims to test models against simulated threats, improve defensive techniques, and share findings with industry partners. The initiative also explores collaboration with governments and organizations to strengthen cybersecurity practices. Project Glasswing reflects growing concern about malicious use of AI and seeks to build protective systems that identify vulnerabilities before attackers exploit them.