Anthropic Teams Up With Its Rivals to Keep AI From Hacking Everything

Anthropic announced the Claude Mythos Preview model and convened Project Glasswing — a consortium including major tech, cybersecurity, and infrastructure organizations — to study and mitigate cybersecurity risks posed by advanced AI. The report highlights that Mythos can discover vulnerabilities, suggest exploit chains, and perform advanced security assessment tasks, stressing the need for coordinated, cautious release and industry preparedness to prevent these capabilities from accelerating malicious activity.