Based on in-depth research on the security risks of large models, 360 Group launched the DS large model security solution, which aims to cover the entire process of large model training, reasoning, and operation, and promote the development of domestic large model enterprises in a safe, virtuous, trustworthy, and controllable direction. The solution provides comprehensive security detection and protection measures for the five major challenges of system security, data security, content security, hallucination problems, and behavioral control. 360 Zhijian is responsible for system security detection, 360 Zhishu protects data security, 360 Zhidun ensures content security, 360 Zhisou reduces the generation of hallucination content, and 360 Zhikong ensures that Agent behavior is controllable. At present, the solution has been applied in government affairs, taxation, medical and other industries, and will continue to provide a "Chinese solution" for global artificial intelligence technology governance in the future.