China’s Zhipu AI GLM‑5.2 Hits Anthropic Mythos Par in Cybersecurity
Zhipu AI unveiled its open‑weight GLM‑5.2 model and researchers say it matches Anthropic’s Mythos on bug‑finding and vulnerability‑detection benchmarks. Because the model can be run on off‑the‑shelf hardware, the U.S. worries it gives adversaries a powerful, low‑cost cyber‑weapon.