GLM-5.3 and the spread of advanced cyber capabilities \ Anthropic

kept by perfcheckacme oct 31

Anthropic's analysis finds GLM-5.3 matches its own Claude Mythos Preview in autonomous cyber exploit development but lacks safeguards, enabling easy misuse by attackers.

The Cloudflare Blog
Cloudflare’s July‑June 2026 blog posts detail advances in cache optimization, post‑quantum crypto, consensus research, AI tooling, monetization, and security initiatives.
also on NIST
Open-weight AI models can exploit the patch gap
Open-weight model GLM 5.3 exploited the Firefox patch gap, turning public security fixes into one-click exploits for six releases at about $1,000 each.
also on GLM-5.3, #cybersecurity, #ai-safety, #open-source
I Asked 100 Agents to Hack Me - by Shrivu Shankar
In a five‑hour test, ~100 abliterated open‑source agents compromised five of the author’s accounts via old‑project vulnerabilities and password brute‑forcing, but failed to breach tier‑0 services.
also on GLM-5.3, #cybersecurity, #open-source
The Flat Curve Society
Steve Yegge argues AI intelligence growth will appear to plateau for most people due to government restrictions and human discernment limits, while actually continuing exponentially behind locked doors.
also on #ai-regulation, #ai-safety, Anthropic, #open-source
Anthropic ditches its core safety promise
Anthropic drops its core safety promise by replacing hard commitments with a flexible, nonbinding framework due to competitive pressure and a shifting political climate.
also on #ai-regulation, #ai-safety, Anthropic