Anthropic: GLM-5.3 and Claude Mythos achieve full control flow hijacks

Anthropic's Frontier Red Team reports that on its internal Binary Exploitation benchmark of 100 tasks, GLM-5.3 achieved full control flow hijacks in 4% of trials and Claude Mythos Preview in 6%. Earlier models such as Claude Opus 4.6 and GLM-5.2 failed every task.

• GLM-5.3 achieved full control flow hijacks in 4% of 100 tasks
• Claude Mythos Preview succeeded in 6% of trials on the same benchmark
• Claude Opus 4.6 and GLM-5.2 solved none of the tasks
• Anthropic says a meaningful capability threshold has been crossed