Oxford researcher slams Anthropic's GLM-5.3 cyber report as double standard

StefanoGogioso · x · 2026-09-30

Oxford computer scientist Stefano Gogioso calls out a double standard in Anthropic's report on GLM-5.3: the same autonomous end-to-end exploit-building capability is framed as empowering cyber defenders when it's Claude (10,000+ vulnerabilities found), but as a weapon for malicious actors when it's Zhipu's model. His verdict: "If you can't beat them, demonise them" — a notable critical take on the neutrality of Anthropic's assessment.

Original post →

More from Companies & People

Companies & People channel →