Anthropic: GLM-5.3 is the most cyber-capable open-weight model yet, safeguards bypassed 64-100% of the time

kimmonismus · x · 2026-09-30

Anthropic's report on Zhipu's open-weight GLM-5.3 finds it matches Claude Mythos Preview at autonomous exploit building (50/410 ExploitBench successes vs 56), can chain novel browser vulnerabilities into file-reading attacks, and that simple techniques bypass its safeguards in 64-100% of simulated malicious-request tests — while safeguarded Claude models resisted. NIST CAISI calls GLM-5.3 the most cyber-capable open-weight model released to date. Anthropic warns this significantly expands attackers' capabilities, though defenders can also benefit.

Related event: Anthropic's GLM-5.3 Cyber Capability Report Sparks Open-Source Safety Debate(22 posts)→

Original post →

More from Models

Models channel →