A Chinese AI model is claimed to rank first for cybersecurity users outside CVP and Daybreak

A post claims the best cybersecurity AI model available to people outside the CVP or Daybreak programs is Chinese, ranking above Mythos, Fable, Astra, Opus and Sol. It says participants still face plenty of classifier downgrades and model rejections. The author argues this could bias open research toward GLM’s capabilities, send more production traces to ZAI and spur more tooling around open models.