人工智能伦理规范核心原则的政策文本内容分析

The Core Framework and Content Analysis of Ethical Principles of Artificial Intelligence

[目的/意义]人工智能技术悄然拉开了第四次工业革命的大幕,为各行各业的发展带来了革命性变化。但该项技术的法律规制和标准规范建设尚不完善,引发了全球对于人工智能伦理原则与规范体系的探索与思考,各国学(协)会相继发布了人工智能伦理规范或政策文件。本研究全面分析和梳理当前认可度较高的人工智能伦理原则的核心框架,旨在为各国制定和完善人工智能伦理规范提供参考。[方法/过程]运用网络调研法、内容分析法和比较研究法,在综合考虑影响力、国别和可获取性等因素的基础上,选取近年各国学(协)会已发布的有关人工智能伦理原则的权威政策文件,提炼出其中涉及的人工智能伦理原则的核心框架和主要内容。[结果/结论]人工智能领域主要包括:造福原则、责任原则、公平正义原则、安全原则、尊重人的自主性原则、隐私与个人信息保护原则、透明性原则和可靠性原则等8项人工智能伦理原则。

[Purpose/significance]Artificial intelligence technology quietly opened the curtain of the fourth industrial revolution,bringing revolutionary changes to the development of all walks of life.However,the legal regulations and standard construction of this technology are not yet perfect,which has triggered the global exploration and thinking of artificial intelligence ethics principles and standard systems.National academic/associations have successively issued artificial intelligence ethics codes or policy documents.This research comprehensively analyzes and sorts out the core framework of the current highly recognized artificial intelligence ethical principles,and aims to provide references for countries to formulate and improve artificial intelligence ethical norms.[Method/process]By using the network research method,content analysis method and comparative research method,on the basis of comprehensive consideration of influence,country,availability and other factors,we select the authoritative policy documents related to the ethical principles of artificial intelligence issued by various national academic associations in recent years,and extract the core framework and main contents of the ethical principles of artificial intelligence involved.[Results/conclusion]The field of artificial intelligence mainly includes 8 artificial intelligence ethical principles,including the benefit principle,the responsibility principle,the fairness and justice principle,the safety principle,the respect for human autonomy,the privacy and protection principle,the transparency principle and the reliability principle.