现代情报2025，Vol.45Issue(12)：16-26,11.DOI:10.3969/j.issn.1008-0821.2025.12.002
基于Sora文生视频深度伪造的情报风险治理机制研究
Research on Intelligence Risk Management Mechanism Based on Sora's Text-to-Video Deep Falsification
摘要
Abstract
[Purpose/Significance]The emergence of Sora's text-to-video has significantly lowered the threshold for video forgery,and deepfake videos may bring a large amount of false intelligence,seriously threatening the security of the intelligence field.Effectively preventing and mitigating risks from Sora's text-to-video deepfake technology in intelligence operations became an urgent imperative.[Method/Process]This research started from the technical principles,analyzed the intelligence risks of text-to-video deepfakes from different dimensions,expounded the governance logic of intelligence risks in the supervision of metaverse sandboxes,and proposed a dynamic closed-loop governance mechanism consisting of"risk prediction and simulation-risk monitoring and intervention-risk review and optimization".[Result/Conclusion]Through the integration of virtual and real scenes,a governance mechanism that combines precision,agility,and resi-lience is used to address the risks of deep forgery of intelligence in Sora's text-to-video,enhance the stability of the intelli-gence ecosystem,and adapt to the new environment of rapid development of artificial intelligence technology,providing new ideas and feasible solutions for intelligence risk governance.关键词
人工智能/文生视频/深度伪造/Sora/情报风险/治理机制Key words
artificial intelligence/text-to-video/deep forgery/Sora/intelligence risk/governance mechanism引用本文复制引用
张涛,张策,崔文波..基于Sora文生视频深度伪造的情报风险治理机制研究[J].现代情报,2025,45(12):16-26,11.基金项目
国家社会科学基金一般项目"数智环境下情报分析算法风险治理路径研究"(项目编号:22BTQ064) （项目编号:22BTQ064）
黑龙江省自然科学基金项目"不确定环境下复杂系统中智能算法可解释模型构建研究"(项目编号:PL2024G019). （项目编号:PL2024G019）
