诊断辅助人工智能应用情境下医疗损害责任认定

The dilemmas and solutions of applying medical malpractice liability to diagnostic-support AI

摘要

诊断辅助人工智能实现了从“规则驱动”到“数据驱动”的技术跃迁，其固有的自主决策性、黑箱性与动态进化性，对以“当时医疗水平”标准为核心的传统医疗损害责任认定构成挑战。对此，应将“诊断辅助人工智能的使用”定位为“新疗法”，以确认其应用的合法性；进而明确该人工智能的输出结果并不直接等同于“当时医疗水平”，且这一标准的内涵亦将动态演进。在此基础上，构建以“审慎再判断义务”为核心的动态过错认定框架，将评价重心转向对医务人员履行审慎使用与结果核验义务过程的审查。这一调整旨在回应当前“人机协作”模式下医疗过错的认定需求，实现技术创新与患者安全保护的规范平衡。

Abstract

Diagnostic-support AI has undergone a technological leap from being rule-driven to data-driven. Its inherent autonomous decision-making capacity，an algorithmic black-box nature，and dynamic evolution pose structural challenges to the traditional medical malpractice liability system，which is centered on the prevailing medical standard. To address these challenges，the use of diagnostic-support AI should be characterized as the adoption of a new medical treatment to establish its legal basis. It is necessary to clarify that its outputs are not directly equivalent to the prevailing medical standard，and that the implications of this standard will also evolve. On this basis，this paper constructs a dynamic fault-determination framework centered on the duty of prudent reassessment. This framework shifts the focus of assessment to evaluate whether healthcare professionals have fulfilled their duties of prudent use and substantive verification of algorithmic outputs，thereby responding to the needs of liability determination in the current human-computer collaboration model and achieving a normative balance between technological innovation and patient safety protection.

关键词

诊断辅助人工智能 / 当时医疗水平 / 医疗过错 / 审慎再判断义务

Key words

diagnostic-support AI / prevailing medical standard / medical fault / duty of prudent reassessment

参考文献

[1] SHORTLIFFE E H, BUCHANAN B G. A model of inex—act reasoning in medicine[J]. Math Biosci, 1975, 23(3/4):351-379

[2] LAGIOIA F, CONTISSA G. The strange case of Dr. Watson：liability implications of AI evidence—based decision support systems in health care[J]. European Journal of Law and Technology（EJLS）, 2020, 12(2):245-289

[3] LIU X H, LIU H, YANG G X, et al. A generalist medical language model for disease diagnosis assistance[J]. Nat Med, 2025, 31(3):932-942

[4] PRICE W N II. Black—box medicine[J]. Harvard Journal of Law & Technology, 2015, 28:419-467

[5] GIUFFRIDA I. Liability for AI decision-making：some legal and ethical considerations[J]. Fordham Law Review, 2019, 88(2):439-456

[6] LIU F, TONG X, YUAN M, et al. Evolution of heuristics：towards efficient automatic algorithm design using large language model［C］. Proceedings of the 41st International Conference on Machine Learning ，Vienna， Austria：PMLR， 2024:1-22

[7] 邹海林, 朱广新. 民法典评注—侵权责任编［M］. 北京： 中国法制出版社， 2020:541-542

[8] VÉLIZ C, PRUNKL C, PHILLIPS—BROWN M, et al. We might be afraid of black—box algorithms[J]. J Med Eth— ics, 2021, 47(5):339-340

[9] 埃里克·托普. 深度医疗［M］. 郑杰，朱烨琳，曾莉娟，译. 郑州： 河南科学技术出版社， 2020:89

[10] LONDON A J. Artificial intelligence and black—box medical decisions：accuracy versus explainability[J]. Hast Cent Rep, 2019, 49(1):15-21

[11] 弗雷德里克·M. 阿尔伯特, 格雷厄姆·杜克斯. 全球医药政策：药品的可持续发展［M］. 翟宏丽，张立新，译. 北京： 中国政法大学出版社， 2016:6-9

[12] 李润生. 论医疗人工智能“黑箱”难题的应对策略和规制进路[J]. 东南大学学报（哲学社会科学版）, 2021, 23(6):83-92，146

[13] COHEN I G. Informed consent and medical artificial intelligence：what to tell the patient?[J]. Georgetown Law Journal, 2020, 108:1425-1469

[14] 郑志峰. 诊疗人工智能的医疗损害责任[J]. 中国法学, 2023(1):203-221

[15] PIERCE R L, VAN BIESEN W, VAN CAUWEN— BERGE D, et al. Explainability in medicine in an era of AI-based clinical decision support systems[J]. Front Genet, 2022, 13:903600

[16] 徐涤宇, 张家勇. 《中华人民共和国民法典》评注［M］. 北京： 中国人民大学出版社， 2022:1288

[17] GIUBILINI A. It is not about AI，it’s about humans. Responsibility gaps and medical AI[J]. J Bioethical Inq, 2025, 22(3):527-537

[18] 赵西巨. 医生对新疗法的使用和告知[J]. 东方法学, 2009(6):119-133

[19] 程啸. 侵权责任法［M］. 3版. 北京： 法律出版社， 2021:644-645

[20] 张新宝. 侵权责任法［M］. 5版. 北京： 中国人民大学出版社， 2020:200

[21] SMITH H, FOTHERINGHAM K. Artificial intelligence in clinical decision—making：rethinking liability[J]. Medi— cal Law International, 2020, 20(2):131-154

[22] PRICE W N Ⅱ. Describing black—box medicine[J]. Bos— ton University Journal of Science & Technology, 2015, 21:347-357

[23] 窦海阳. 法院对医务人员过失判断依据之辨析——以《侵权责任法》施行以来相关判决为主要考察对象[J]. 现代法学, 2015, 37(2):167-181

[24] PRICE W N II, GERKE S, COHEN I G. Potential liability for physicians using artificial intelligence[J]. JAMA, 2019, 322(18):1765-1766

[25] 刘建利. 医疗人工智能临床应用的法律挑战及应对[J]. 东方法学, 2019(5):133-139

[26] GERKE S, MINSSEN T, COHEN G. Ethical and legal challenges of artificial intelligence—driven healthcare［A］// Artificial Intelligence in Healthcare. Amsterdam：Elsevier， 2020:295-336

[27] 徐辉, 王译. “硬法—软法”范式下医疗人工智能伦理治理路径探析[J]. 南京医科大学学报（社会科学版）, 2024, 24(4):359-368

[28] 黄薇. 中华人民共和国民法典侵权责任编解读［M］. 北京： 中国法制出版社， 2020:210

[29] 赵丹娜, 赵蓝蓝, 陈任, 等. 某三级公立医院医务人员对人工智能医学应用伦理问题的认知研究[J]. 南京医科大学学报（社会科学版）, 2022, 22(5):465-470

[30] 孙运梁, 王璨. 医疗人工智能介入过失诊疗行为的结果归属研究[J]. 江西师范大学学报（哲学社会科学版）, 2025, 58(5):137-148

[31] MAROUDAS V P. Fault—based liability for medical malpractice in the age of artificial intelligence：Α comparative analysis of German and Greek medical liability law in view of the challenges posed by AI systems[J]. Rev Eur Comp Law, 2024, 57(2):135-169

基金资助

中共浙江省委政法委员会浙江省法学会法学研究课题重点项目“医疗人工智能法律规制与侵权责任归属研究”(2025NA32)