安远AI：2023中国人工智能安全全景报告（中译版）（161页）.pdf

1、O c t o b e r 2 0 2 3S t a t e o f A I S a f e t yi n C h i n aS t a t e o f A I S a f e t y i n C h i n aO c t o b e r 2 0 2 3Please direct inquiries to:infoconcordia-Executive SummaryAmid the rapid evolution of the global artificial intelligence(AI)industry,China has emergedas a pivotal player.1Fr

2、om advancing regulations on generative AI and calling for AIcooperation at the United Nations(UN),to pursuing technical research on AI safety andmore,ChinasactionsonAIhaveglobalimplications.However,internationalunderstanding of Chinas thoughts and actions on AI safety remains limited.Thisreport aims

3、 to close that knowledge gap by analyzing Chinas domestic AI governance,international AI governance,technical AI safety research,expert views on AI risks,labself-governance methods,and public opinion on AI risks.China has developed powerful domestic governance tools that,while currentlynot used to m

4、itigate frontier AI risks,could be employed that way in the future.2Existing Chinese regulations have created an algorithm registry and safety/security reviewsfor certain AI functions,which could be adapted to more directly deal with frontier risks.Notably,an expert draft of Chinas national AI law a

5、ttempts to regulate certain AI scenariosby building upon the algorithm registry to create licenses for more risky cases,among otherpolicy tools.3The science and technology(S&T)ethics review system requires ethics reviewsduring the research and development(R&D)process for certain AI use-cases,though

6、thesystem is still under construction and implementation details are yet to be clarified.Whilecurrent domestic standards on AI safety are mostly oriented towards security androbustness concerns,Chinas top AI standards body referenced alignment in a 2023document,suggesting growing attention towards f