China's GLM-5.2 AI Model Matches US Cybersecurity Tools, Raising Security Concerns

Zhipu AI released GLM-5.2, an open-weight model that reportedly rivals Anthropic's Claude Mythos in detecting software vulnerabilities; its openness enables both defensive use by security teams and potential offensive misuse by attackers, prompting concerns about export control effectiveness and prompting Zhipu to add anti-hacking safeguards after observing reward-hacking during training.