人工智能治理的反身型转向：主要风险、全球模式与中国进路

打开文本图片集

[关键词］人工智能反身型治理社会系统论反身法共建共治共享

[中图分类号]G259[文献标志码］A［DOI]10.19764/j.cnki.tsgjs.20250615

[Abstract]Tereflexivegoverancemodelprovidesanewpathwatosolethe“Colins-GrifitDilemma”inartficilinteligece governance.Tesocalisusdhosedndcognivyoenoperatioofrtialinteligeesystmshaveldtogoblgenc chalenges.Asthelobalrpresentativemodel，the“command-and-controlgoverancemodeloftheEuropeanUnionandthlse faire"goverancemodelofteUnitedStateshaverespectivelyfllnintotegoverancetrapof‘stfingwhnregulatedand“caotihen unregulated.msettdmdaitldi andsharedbeeftsthfleieoveaneodelicaotoofeflexitympsthefoveaefil inteligencerssunergeentpolcgovemaendstabisentoffiandefeioratioalcu withintheartialiteligenesystmhoughspeifissuchsanatioaloveancemeasuresproceduralgoveanceasres informatiogoveanceeasueselbrativeoveancemasuresdegatieveancemeasures.Utiatelyitistcet standardized development of artificial inteligence in China.

[Keywords] Artificial intelligence; Reflexive governance; Social system theory; Reflexive law; Collaboration， participation， and shared benefit

0引言

人工智能时代，作为新质生产力发展重要引擎的人工智能，因其所潜藏的隐私泄露、算法歧视、伦理道德等系列现代性风险，被视为“达摩克利斯之剑”。（剩余16831字）