GLM-5.3 Released with Advanced Cyber Capabilities, Reportedly Finds 'Serious Vulnerability' in Cursor

Summary

Chinese AI startup Z.ai has released GLM-5.3, an advanced language model with significant improvements in long-horizon coding and cybersecurity capabilities, reportedly uncovering a serious vulnerability in Cursor. The model, which builds on GLM-5.2's base model through scaled post-training, is initially available via Z.ai's Coding Plan and ZCode environment, with API access and open weights to follow after safety evaluations.