情感计算的应用困境及其法律规制  

The Application Dilemma and Legal Regulation of Affective Computing

作　　者：王禄生[1,2] Wang Lusheng

机构地区：[1]东南大学法学院 [2]东南大学人民法院司法大数据研究基地

出　　处：《东方法学》2021年第4期49-60,共12页Oriental Law

基　　金：国家重点研发计划项目(项目批准号:2018YFC0830200);东南大学中央高校基本科研业务费专项资金项目(项目批准号:2242021K10011)的阶段性研究成果。

摘　　要：在人工智能的视域下,情感计算是感知、识别、模拟、影响人类情感状态的各类智能技术的统称。尽管该技术已经在医疗、教育、商业、就业、执法等领域获得应用,但仍然面临法律、伦理与技术的多维困境。从法律上看,情感计算的应用可能与个人信息权、隐私权、人身自由权等形成冲突;从伦理上看,情感计算的终极目标与道德哲学领域对情感自主性的强调不相兼容;从技术上看,情感计算仍然面临着基础理论与实际成效的质疑。为了推动情感计算应用的良性发展,需要从两方面展开法律规制:一是以"轻推"为限度,调和情感自主性与情感计算终极目标之间的张力;二是以风险为基础,按照比例原则对不可接受、高风险、中风险和低风险四类应用实施从完全禁止到行业自律的不同规制。In the perspective of artificial intelligence,affective computing is a collective term for all kinds of intelligent technologies that perceive,emotion recognize,simulate and influence human affect.Although this technology has been applied in medical treatments,education,business,employment,law enforcement and other fields,it still faces multidimensional dilemmas in law,ethics and technology.In terms of law,the application of affective computing may be in conflict with the right to personal information,privacy and personal freedom.In terms of ethics,the ultimate goal of affective computing is incompatible with the emphasis on affect autonomy in the field of morality philosophy.In terms of technology,affective computing is still facing the doubts of the basic theory and practical effect.In order to promote the sound development of the application of affective computing,it is necessary to carry out legal regulation from two aspects:The first is to mediate the tension between affect autonomy and the ultimate goal of affective computing with"nudging"as the limit;Secondly,based on risks,different regulations should be implemented in accordance with the principle of proportionality,ranging from total prohibition to industry self-discipline for the four categories of applications of inadmissible risk,high risk,moderate risk and low risk.

关 键 词：情感计算 情绪识别 情绪智能 人工智能 比例原则 法律规制 

分 类 号：DF0[政治法律—法学理论]