China's Zhipu AI has released its open-weight GLM-5.2 model, which some researchers claim can match Mythos in certain bug-finding and cybersecurity scenarios.
While GLM-5.2 lags behind models from Anthropic and OpenAI in general tasks, it has dramatically reduced the gap in capabilities between Chinese and US models.
This advancement is concerning to the US government, which has restricted China's access to powerful models and hardware. The open-weight nature of GLM-5.2 allows it to be downloaded and run by anyone, making it flexible but also vulnerable to abuse by bad actors.
