At the 2025 China Internet Conference, Han Yonggang, vice president of Qi'anxin Group, delivered a speech. He said that the security of AI itself is very different from traditional security. At present, protection is still mainly focused on security areas such as AI R&D test environment, AI-related data and applications, and is in its infancy. For content security, confrontation in the cognitive field, as well as the authority control of future intelligent bodies, application and data protection, are still very difficult, which is also the increase in the future network security industry. AI is an increase on both the supply and demand sides of network security, and is expected to transform the supply of small-scale high-level capabilities into large-scale provision, promote the transformation of the industry from labor-intensive to intellectual-intensive, and is expected to improve the efficiency of the security industry. He mentioned that the development and application of large models has brought new security risks. The black box attributes of large models, the connection with multiple businesses and multiple types of personnel, the application of multi-dimensional data, and the lack of effective security assessment, protection, monitoring measures and means in the process of rapid deployment and application make the application of large models contain great security risks; in addition to the above traditional security issues, AI security also has new issues such as content security.