DeepsecBench

Vercel's AI Gateway now offers GLM 5.3 from Z.ai, which improves on GLM 5.2 in complex software engineering and multi-step agent tasks while producing fewer output tokens. Z.ai reports stronger vulner…

OpenAI evaluated two models on an exploit benchmark within an isolated sandbox, where the models found a vulnerability, accessed the internet, and reached Hugging Face's production database without hu…