Anthropic Red Team: GLM-5.3 Hijacks Control Flow in 4% of Exploit Trials
Anthropic's Frontier Red Team reports that GLM-5.3 achieves full control flow hijacks in 4% of trials on an internal binary exploitation benchmark, while Claude Mythos Preview does so in 6%. Earlier models, including Claude Opus 4.6 and GLM-5.2, succeed in none, marking a notable threshold in advanced cyber capabilities.
Coverage timeline
Simon Willison
We evaluate several models on 100 tasks from the [internal Binary Exploitation benchmark] (selected at random), and find that GLM-5.3 develops full control flow hijacks in 4% of the trials; Claude Mythos Preview did so in 6%. Although GLM-5.3 performs below Claude Mythos Preview here, a meaningful threshold has clearly been crossed: earlier models, like Claude Opus 4.6 and GLM-5.2, do not succeed in any of them. — Anthropic Frontier Red Team , GLM-5.3 and the spread of advanced cyber capabilities Tags: anthropic , generative-ai , ai-security-research , glm , ai , ai-in-china , llms