China's "Mythos-Class" AI Closer: Zai Chief on GLM 5.2

A Chinese AI lab suggests China could develop a frontier AI model comparable to today's most advanced systems sooner than expected. This follows the launch of GLM 5.2, an open-weight model trained entirely on Huawei chips. Tang Jie, founder of Chinese AI lab Zai, made these comments. He disagreed with predictions that such a development would take until late 2026 or early 2027. What's interesting is GLM 5.2's performance. It achieved a score of 51 on the Artificial Analysis Intelligence Index, surpassing models like Google's Gemini 3.5 Flash and Anthropic's Claude Sonnet 4.6. However, it still trails newer systems like Anthropic's Fable 5 and OpenAI's GPT 5.5. The model was trained using Huawei Ascend chips, not Nvidia hardware. This is significant because US export restrictions limit China's access to advanced Nvidia processors. The success of GLM 5.2 suggests Chinese companies are finding alternative paths despite these restrictions. The bottom line is this shows the intensity of the global AI competition.