Workflow
奇安信董事长齐向东:人工智能大模型爆发 引发三重安全风险

Group 1 - The application of large AI models is experiencing a significant explosion across various industries, bringing both immense productivity and creativity, as well as substantial security risks [2] - "Small data" has become a core concern in the digital economy, where data is likened to "oil" and AI to an "engine." Traditional enterprises are now processing their accumulated data into refined "small data," which, if compromised, can lead to a collapse of their core competitiveness [2][3] - There is a high dependency on AI by some enterprises, which can lead to a "butterfly effect" if large model decisions go awry. External attacks can exploit vulnerabilities, while internal errors can contaminate the model's learning environment [3] Group 2 - The majority of active servers for large models are operating without proper security measures, allowing unauthorized access and potentially leading to a chain reaction of business disruptions [3] - The importance of network security is emphasized as a foundational element for the development of the digital economy, with the need for continuous innovation in cybersecurity to create a secure environment for new technologies [4] - The global technological competition is intensifying, with AI at the center, necessitating advancements in cybersecurity to ensure technological autonomy and create a protective barrier for the digital economy [4]