AI Cyber Model Raises Security Fears

Anthropic’s advanced AI system “Claude Mythos Preview” is triggering serious cybersecurity concerns across governments, regulators and major technology firms worldwide. The unreleased model reportedly demonstrated the ability to discover critical vulnerabilities in operating systems, browsers and core internet infrastructure: capabilities experts warn could redefine the future of cyber warfare and digital defence. While Anthropic says the project is designed to strengthen cybersecurity through “Project Glasswing,” fears are growing over misuse, AI-driven exploits and a widening global cyber gap.