摘要
针对生成式AI发展要求,基于算法固有的“黑箱”问题,分析了当前生成式AI算法安全所面临的风险,并从健全算法透明体系、加强算法伦理建设、完善算法问责机制等三个方面提出了加强算法安全管理的对策建议。
In view of the development requirements of generative Al,based on the inherent"black box"problem of algorithms,we analyze the current risks faced by algorithm security of generative Al,and propose countermeasures to strengthen algorithm security management in three aspects,including improving algorithm transparency system,strengthening algorithm ethics construction,and improving algorithm accountability mechanism.
作者
高俊
GAO Jun(College of Law,Anhui University,Hefei 230039,China)
出处
《通信与信息技术》
2023年第4期122-124,128,共4页
Communication & Information Technology
