刑罚的有效性与人工智能领域的刑事责任主体
|
摘要
为有效预防人工智能给社会带来的风险,将之纳入刑法调整的对象是必要的。因此,应将谁作为调整对象就成为必须予以回答的问题。有的观点从法人刑事责任、人工智能的道义性等角度出发,主张将人工智能及其载体作为刑事责任的主体。但是,人工智能是以运算逻辑与数据为核心的计算机程序,不具有道义性,与由具备道义选择能力的个人组成的法人也存在本质区别,而且将人工智能或者其载体作为刑事责任主体不但可能导致人工智能行为的滥用,而且会使现有刑罚体系失效。为充分发挥刑罚的预防功能,有效预防人工智能可能带来的风险,立法机关应将研发者、生产者以及使用者规定为刑事责任主体,并通过采纳严格责任、明确义务边界等立法措施,避免对技术进步造成障碍。
In order to effectively regulate the risks that artificial intelligence brings to society,it is necessary to include it in the adjustment of criminal law. Then,who should be the subject of adjustment becomes a crucial question that must be answered. Some viewpoints advocate the AI and AI agent as the subject of criminal responsibility from the perspective of legal person's criminal responsibility and the morality of artificial intelligence. However,artificial intelligence is a computer program with computing logic and data as its core. It is not moral,and it is also fundamentally different from a legal person composed of individuals with moral choice ability. With AI or AI agent as the subject of criminal responsibility,it is possible to cause the abuse of artificial intelligence behavior and invalidate the existing penalty system. In order to give full play to the preventive function of penalties and effectively prevent the risks that artificial intelligence may bring,the legislature should stipulate the developers,producers and users as criminal liability subjects,and avoid impeding obstacles to technological progress,by adopting legislative measures such as strict liability and clear obligation boundaries.
In order to effectively regulate the risks that artificial intelligence brings to society,it is necessary to include it in the adjustment of criminal law. Then,who should be the subject of adjustment becomes a crucial question that must be answered. Some viewpoints advocate the AI and AI agent as the subject of criminal responsibility from the perspective of legal person's criminal responsibility and the morality of artificial intelligence. However,artificial intelligence is a computer program with computing logic and data as its core. It is not moral,and it is also fundamentally different from a legal person composed of individuals with moral choice ability. With AI or AI agent as the subject of criminal responsibility,it is possible to cause the abuse of artificial intelligence behavior and invalidate the existing penalty system. In order to give full play to the preventive function of penalties and effectively prevent the risks that artificial intelligence may bring,the legislature should stipulate the developers,producers and users as criminal liability subjects,and avoid impeding obstacles to technological progress,by adopting legislative measures such as strict liability and clear obligation boundaries.
引文
[1]叶良芳,马路瑶.风险社会视阈下人工智能犯罪的刑法应对[J].浙江学刊,2018(6):65-72.
[2]刘宪权.涉人工智能犯罪刑法规制的路径[J].现代法学,2019(1):75-83.
[3]蔡婷婷.人工智能环境下刑法的完善与适用[J].犯罪研究,2018(2):21-26.
[4]皮埃尔·斯加鲁菲.离人工智能奇点还有多远[J].王艺璇,译.中国经济报告,2017(5):112-113.
[5]Dennis Coon等.心理学导论[M].郑钢,等,译.北京:中国轻工业出版社,2014:187.
[6]王燕玲.人工智能时代的刑法问题与应对思路[J].政治与法律,2019(1):22-34.
[7]刘宪权.人工智能时代的刑事风险与刑法应对[J].法商研究,2018(1):3-11.
[8]谢冶东.论单位法人刑事责任之本质[J].湖北社会科学,2011(10):133-137.
[9]马骏驹,刘卉.论法律人格内涵的变迁和人格权的发展[J].法学评论,2002(1):26-41.
[10]储陈城.人工智能时代刑法归责的走向[J].东方法学,2018(3):27-37.
[11]郝铁川.不可幻想和高估人工智能对法治的影响[N].法制日报,2018-01-03(3).
[12]王世洲.现代刑罚目的与中国的选择[J].法学研究,2003(3):107-131.
[13]邱兴隆.报应刑的价值悖论——以社会秩序、正义与个人自由为视角[J].政法论坛,2001(2):21-30.
(1)该概念由美国哲学家约翰·塞尔于20世纪七十年代提出,主要内容为基于心智的计算模型,以通用数字计算机为载体的AI程序可以像人类一样认知和思考,达到或者超过人类智能水平。参见陈自富:《强人工智能与超级智能:技术合理性及其批判》,载《科学与文化》2016年第5期。
(1)该理论认为,我们即将迎来机器时代:机器的智能程度远远超过人类,以至于人类既无法控制机器,也无法理解它们的想法。参见皮埃尔·斯加鲁菲:《离人工智能奇点还有多远》,王艺璇译,载《中国经济报告》2017年第5期。
(1)刑法人类中心主义是指,以人类中心主义思维模式构建的可以限制刑法修正案犯罪化的各种原理、学说或原则,希望通过人类某种共享的价值系统进行意义沟通,从而实现刑法的演化并构建社会所需之秩序。参见姜敏:《系统论视角下刑法修正案犯罪化限制及其根据》,载《比较法研究》2017年第3期。
[2]刘宪权.涉人工智能犯罪刑法规制的路径[J].现代法学,2019(1):75-83.
[3]蔡婷婷.人工智能环境下刑法的完善与适用[J].犯罪研究,2018(2):21-26.
[4]皮埃尔·斯加鲁菲.离人工智能奇点还有多远[J].王艺璇,译.中国经济报告,2017(5):112-113.
[5]Dennis Coon等.心理学导论[M].郑钢,等,译.北京:中国轻工业出版社,2014:187.
[6]王燕玲.人工智能时代的刑法问题与应对思路[J].政治与法律,2019(1):22-34.
[7]刘宪权.人工智能时代的刑事风险与刑法应对[J].法商研究,2018(1):3-11.
[8]谢冶东.论单位法人刑事责任之本质[J].湖北社会科学,2011(10):133-137.
[9]马骏驹,刘卉.论法律人格内涵的变迁和人格权的发展[J].法学评论,2002(1):26-41.
[10]储陈城.人工智能时代刑法归责的走向[J].东方法学,2018(3):27-37.
[11]郝铁川.不可幻想和高估人工智能对法治的影响[N].法制日报,2018-01-03(3).
[12]王世洲.现代刑罚目的与中国的选择[J].法学研究,2003(3):107-131.
[13]邱兴隆.报应刑的价值悖论——以社会秩序、正义与个人自由为视角[J].政法论坛,2001(2):21-30.
(1)该概念由美国哲学家约翰·塞尔于20世纪七十年代提出,主要内容为基于心智的计算模型,以通用数字计算机为载体的AI程序可以像人类一样认知和思考,达到或者超过人类智能水平。参见陈自富:《强人工智能与超级智能:技术合理性及其批判》,载《科学与文化》2016年第5期。
(1)该理论认为,我们即将迎来机器时代:机器的智能程度远远超过人类,以至于人类既无法控制机器,也无法理解它们的想法。参见皮埃尔·斯加鲁菲:《离人工智能奇点还有多远》,王艺璇译,载《中国经济报告》2017年第5期。
(1)刑法人类中心主义是指,以人类中心主义思维模式构建的可以限制刑法修正案犯罪化的各种原理、学说或原则,希望通过人类某种共享的价值系统进行意义沟通,从而实现刑法的演化并构建社会所需之秩序。参见姜敏:《系统论视角下刑法修正案犯罪化限制及其根据》,载《比较法研究》2017年第3期。