According to Dynamic Beating monitoring, Anthropic has confirmed in the latest release of its flagship model, Claude Opus 4.8, that the development team has made a breakthrough in security defense and protection mechanisms. It is expected to roll out the next-generation Mythos-class model with super flagship intelligence to all customers in the coming weeks. In the Glasswing software security project report released on May 22nd, the research team emphasized that the Mythos-class model would not be made available to the public until more effective security measures are developed.
As a cutting-edge model with end-to-end penetration testing capabilities and other highly advanced network defense strengths, the public release of the Mythos-class model has long been hindered by the significant gap between "vulnerabilities discovered in seconds, patches deployed in weeks." To prevent misuse of the model and mitigate potential harm, the Anthropic team is currently accelerating the development of more robust security defenses. Additionally, the team has started providing scanning orchestration tools to eligible security teams to help them establish an asymmetric advantage in defense.
