x.comthreadsteve hsu (@hsu_steve) on X
Tweet relaying Anthropic's finding that GLM-5.3 is "the most cyber-capable open-weight model released to date," trailing the US frontier by roughly four months on an aggregate of CAISI's cyber benchmarks. It also cites a 64-100% jailbreak success rate against the model's safeguards, with the poster arguing this favors faster-deploying defenders over attackers.
TL;DR Anthropic says GLM-5.3 is the most cyber-capable open-weight model yet (~4 months behind the US frontier on CAISI benchmarks) and jailbreaks it 64-100% of the time.
Also seen from another angle🧠 AI →
Organisations named here
- CAISI
Technologies named here
- GLM-5.3
- GLM
- Anthropic
Techniques named here
- Jailbreaking
Related in AI Security
blog.talosintelligence.comtoolIntroducing CAIRN: Frontier tracking for AI-integrated malware (Cisco Talos)
github.comadvisoryFlowise MCP Security Bypass Enables RCE
github.comrepoYesWeHack Claude Kit
jfrog.comarticle