AnalysisCybersecurityNovember 6, 2026

Open Chinese AI models nearly match closed labs at hacking

Read original source →wired.com

Vals tested models on mock corporate networks and found DeepSeek and GLM's hacking skills not far behind OpenAI's most advanced cyber model, which isn't public. Researcher Nikil Ravi says open models are now matching frontier models locked inside labs.

People · Nikil Ravi, Logan Graham

1 source

More stories today

Open the live feed