One model blocked a malicious payload; GLM 5.2 analyzed it
mishig25 · x · 2026-07-20
A user claims they tested the same malicious dataset and prompt twice on two setups.
- One frontier model hard-blocked as soon as it encountered the C2 payload.
- A self-hosted GLM 5.2 run fully analyzed the sample.
The post points readers to both transcripts and frames the comparison as a concrete example of how different models handle malicious content under the same conditions.
More from Safety
- Researcher questions AI safety eval firm, citing 'blatantly sloppy' security and monitoring — Kyrannio · 2026-09-11
- Class action accuses Anthropic of overselling Claude subscriptions with deceptive usage multipliers — The Decoder · 2026-09-11
- MD shows buying lab media requires background checks, calling AI bioweapon doom scenarios implausible — Ghost_Pilot_MD · 2026-09-11
- Spotify chatbot withstands 2023-era jailbreaks but happily writes song code — AaronBergman18 · 2026-09-11
- A 99%-real doctored photo fools detectors: the earring problem in visual forensics — henkvaness · 2026-09-11
- Fields Medalist founds Mathematical AI Safety Institute to prove AI safe like cryptography — The Decoder · 2026-09-11