Refactoring ai risk governance:from inside perspective to outside perspective
The academic research on artificial intelligence risk governance focuses on the artificial intelligence compliance system based on the internal perspective,which is unable to fully reveal the institutional core of artificial intelligence risk governance.From the external perspective of risk governance,if AI compliance is placed under the framework of government-enterprise relationship,the basic concept of AI compliance will no longer be pure self-regulation,but the interaction between self-regulation and government regulation.Under such a change of perspective,a"outsourcing"relationship is formed between platform enterprises and the government on AI governance task assignment,which is reflected as a platform outsourcing system and can be specifically expanded into four sub-systems:authorization mechanism is the institutional premise of platform outsourcing system,that is,the government grants regulatory power to platform enterprises based on trust,and such authorization is certified through privacy policies;The evaluation mechanism is used to measure the artificial intelligence risk management ability of the platform enterprises;Both incentive mechanism and assessment mechanism aim to improve the AI risk governance ability of platform enterprises.The former promotes platform enterprises to participate in the competition based on AI risk governance ability through tournament theory,while the latter promotes platform enterprises to form indicators reflecting AI risk governance ability through pyramid theory.
artificial intelligence risk governanceinternal perspectiveexternal perspectiveAI complianceplatform contract system
NETL
NSTL
万方数据
