摘要
人工智能技术迅猛发展,在为经济建设持续注入新动能的同时,所伴生的系统性风险不断产生外溢效应,危及个人、社会和国家安全,导致加强人工智能安全监管成为应对人工智能安全风险的必然选择。当前,各国对此采取了不同监管模式,总体看还存在监管时间滞后、空间分化、责任模糊和知识欠缺四大困境,亟须加强系统性研究,重塑监管模式。在生态学和系统论视域下,可以运用与人工智能安全监管具有高度适配性的共生理论来构建“共生型”监管模式,即从政府单向线性管控转向多元网状协同监管,从静态安全合规审查转向共生型动态风险承担,从零和博弈的国家竞争转向合作共赢的全球共治,通过完善各主体“权责利共生”的监管生态体系,有效解决当前监管困境,更好地统筹人工智能高质量发展和高水平安全,开创“人—机—人”互惠共生的可持续未来。
The rapid advancement of artificial intelligence(AI)technology has continuously injected new momentum into economic development.However,the systemic risks that come with such progress have been generating spillover effects.threatening the security of individuals,society,and nations.Consequently,strengthening AI safety oversight has become an inevitable choice for addressing AI safety risks.Currently,countries have adopted different oversight approaches,but on the whole,four major problems persist:oversight time lag,spatial fragmentation,ambiguous responsibilities,and knowledge gaps.There is an urgent need to strengthen systematic research and reshape oversight models.From the perspectives of ecology and systems theory,the theory of symbiosis—highly compatible with AI safety oversight—can be employed to construct a"symbiotic"oversight model.This entails shifting from unilateral,linear government control to multi-stakeholder networked collaborative oversight,and transitioning from static compliance reviews to symbiotic dynamic risk-sharing,and from zero-sum competition to cooperative win-win global governance.By refining a oversight ecosystem where all stakeholders share symbiotic rights,responsibilities,and benefits,the symbiotic model effectively resolves current oversight problems,and better balances the highquality development of AI with high-level safety,paving the way for a sustainable future of mutual benefit and symbiosis among humans,machines,and humans.
出处
《国家安全论坛》
2025年第5期41-55,96,共16页
National Security Forum
基金
2025年度上海市教育科学研究项目“新时代青少年人工智能安全教育协同机制创新研究”(项目号:C2025106)的阶段性成果。
