BLAST RADIUS

Open-weight & owned stack · Benchmarks evals

Anthropic Reports GLM-5.3 Can Autonomously Build Cyber Exploits

Sep 30, 2026, first seen via OpenAI & Anthropic publications. Covered by 7 outlets.

Anthropic reported that Z.ai’s open-weight GLM-5.3 chained browser vulnerabilities and performed cyber tasks near the capability of restricted frontier models while often bypassing safeguards.

Read the original at anthropic.comOpens the publisher's site in a new tab

Also covering this

More in Open-weight & owned stack