Anthropic: GLM-5.3 safeguards bypassed 64%-100% of the time in cyber exploit tests
AlexTensor · x · 2026-09-30
Anthropic published an analysis of GLM-5.3, Zhipu AI's latest model, assessing the spread of advanced cyber capabilities. Key findings:
- Like Anthropic's own Claude Mythos Preview, GLM-5.3 can autonomously build end-to-end cyber exploits—but was released without meaningful misuse safeguards.
- In Anthropic's simulated tests, attackers bypassed GLM-5.3's safeguards 64%–100% of the time with simple techniques; the same attacks failed against safeguarded Claude models.
- NIST's CAISI independently called GLM-5.3 "the most cyber-capable open-weight model released to date."
Context: five months earlier Anthropic had released Claude Mythos Preview in limited form via Project Glasswing, which enabled trusted defenders to find 10,000+ vulnerabilities in critical software. Anthropic assesses GLM-5.3's lax safeguards significantly expand offensive cyber capabilities available to malicious actors, while noting defenders can also benefit.
More from Models
- Fable 5.2 Leads MazeBench Token Efficiency; Billion-Token Runs Underway — patience_cave · 2026-11-16
- Fable 5.2's Puzzle Skills Look Suspiciously Good, Reciting Outdated MazeBench Solution — patience_cave · 2026-11-16
- Fable 5.2 Tops MazeBench With 21% After 90 Hours and $6,000 in Credits — patience_cave · 2026-11-16
- Tencent Hy4 flash and InternLM Sage-S2.1-Flash both released, tool-calling tests requested — MajesticPuzzler · 2026-11-16
- User Shares Coral Reef Creation Labeled 'Qwen 4 Max (xhigh)', an Unannounced Model Name — intellectronica · 2026-11-16
- Stealth Model 'silicon-2' Surfaces in Devin, Suspected to Be Anthropic's Next Model — kimmonismus · 2026-11-16