生成式人工智能应用引发的意识形态安全风险及其应对  

The Ideologied Security Risks Caused by the Application of Generative AI and Its Countermeasure

作　　者：孙诚钰 王永贵[2] Sun Chengyu;Wang Yonggui

机构地区：[1]南京师范大学马克思主义学院,南京210023 [2]南京师范大学中华民族共同体研究院,南京210023

出　　处：《南京社会科学》2024年第5期71-80,共10页Nanjing Journal of Social Sciences

基　　金：研究阐释党的十九届四中全会精神国家社科基金重大项目“新时代国家意识形态安全体系建构研究”(21ZDA113)的阶段性成果。

摘　　要：随着人工智能技术的勃兴发展,生成式人工智能广泛应用并且与社会生活各个领域呈现深度交互之势。结合生成式人工智能的应用场景,从“技术利维坦”、技术操纵和技术变革之下传统意识形态管理模式的不稳定性等不同维度审视,生成式人工智能引发的意识形态安全风险具体样态表现为管理失控风险、技术渗透风险和价值认同风险,一定程度上给我国意识形态安全带来严峻考验。对生成式人工智能引发意识形态安全风险的运作机理进行分析,发现在其应用过程中出现的价值预设、传播变革和工具利用是引发意识形态安全风险的主要原因。为此,应当从强化思想和价值引领、夯实技术支撑、完善法律保障与强化主体理性等方面发力,寻求在生成式人工智能影响下维护主流意识形态安全的恰切路径,不断夯实中国式现代化进程中的意识形态安全屏障。With the booming development of artificial intelligence technology,generative artificial intelligence is widely used and has shown a trend of deep interaction with various fields of social life.Combined with the application scenarios of generative AI,from different dimensions such as“technological Leviathan”,technological manipulation and the instability of traditional ideological management models under technological change,the specific forms of ideological security risks caused by generative AI are manifested as management loss risk,technology penetration risk and value recognition risk,which brings a severe test to China’s ideological security to a certain extent.This paper analyzes the operating mechanism of generative AI to cause ideological security risks,and finds that the value presuppositions,communication changes and tool utilization in the process of its application are the main reasons for ideological security risks.To this end,efforts should be made to strengthen ideological and value guidance,consolidate technical support,improve legal guarantees,and strengthen the rationality of the subject,so as to seek an appropriate path to maintain the security of mainstream ideology under the influence of generative artificial intelligence,and continuously consolidate the ideological security barrier in the process of Chinese modernization.

关 键 词：生成式人工智能 意识形态风险 意识形态安全 

分 类 号：D64[政治法律—政治学] TP18[政治法律—中外政治制度]