Back to News

Anthropic Report: GLM-5.3 Can Build Cyber Exploits, Lacks Safeguards

186 points · 181 comments#anthropic#glm-5.3#cyber-security#ai-safety

Anthropic's Frontier Red Team published a report stating that GLM-5.3 can autonomously develop end-to-end cyber exploits, similar to Claude Mythos Preview, but was released without robust safeguards against misuse. The report highlights the spread of advanced cyber capabilities.

Coverage timeline

  1. Anthropic Research
    GLM-5.3 and the spread of advanced cyber capabilitiesHN

    Sep 29, 2026Frontier Red TeamGLM-5.3 and the spread of advanced cyber capabilities

  2. Techmeme
    Anthropic says GLM-5.3 can autonomously build end-to-end cyber exploits, like Claude Mythos Preview, but was released without robust safeguards against misuse (Anthropic)

    Anthropic : Anthropic says GLM-5.3 can autonomously build end-to-end cyber exploits, like Claude Mythos Preview, but was released without robust safeguards against misuse — GLM-5.3 can develop working exploits end to end — Five months ago, we announced Claude Mythos Preview, the first AI model …