• 中国核心期刊数据库收录期刊
  • 中文科技期刊数据库收录期刊
  • 中国期刊全文数据库收录期刊
  • 中国学术期刊综合评价数据库统计源期刊等

快速检索引用检索图表检索高级检索

监管科学

药品监管人工智能伦理治理初探

  • 陈锋 ,
  • 由玉伟
展开
  • 国家药品监督管理局信息中心,北京 100076
陈锋,男,研究员,研究方向:药品智慧监管。

收稿日期: 2025-09-25

  修回日期: 2025-12-22

  录用日期: 2025-12-24

  网络出版日期: 2025-12-24

Exploring Ethical Governance of Artificial Intelligence in Drug Regulation

Expand
  • Center for Information of NMPA Beijing 100076, China

Received date: 2025-09-25

  Revised date: 2025-12-22

  Accepted date: 2025-12-24

  Online published: 2025-12-24

摘要

近年来,以生成式人工智能为代表的新技术迅猛发展,深刻重塑着各行各业的基本范式。在药品监管这一关乎公众生命健康的特殊领域,人工智能技术的引入在提升审评审批、日常监管、服务公众与辅助决策效能的同时,也因其技术的复杂性与不确定性带来了独特的伦理挑战。本研究系统梳理了人工智能在药品监管中的典型应用场景,并深入剖析了其可能引发的内生风险(如算法偏见与内容幻觉)、应用风险(如模型安全漏洞与数据隐私泄露)以及衍生风险(如责任主体模糊与监管公信力受损)。通过借鉴世界卫生组织、联合国教科文组织、欧盟等主要国际组织与其他监管机构的治理经验,结合我国实际,构建了覆盖“理念-规范-治理-能力”4个层次的综合性人工智能伦理治理体系。该体系主张在理念层确立以人民健康为中心、发展与安全并重的价值导向;在规范层推动科技伦理与医学伦理要求的协同融合,并制定专门的技术标准;在治理层建立基于风险的全生命周期监管机制,确保可追溯性与问责制;在能力层着力打造复合型人才队伍与技术支持平台。通过该伦理体系的建立与完善,为推动人工智能在我国药品监管领域实现稳健、负责任的应用提供理论支撑与实践框架,从而切实保障公众用药安全与健康权益。


本文引用格式

陈锋 , 由玉伟 .

药品监管人工智能伦理治理初探

[J]. 中国医药导刊, 2025 , 27(11) : 1121 -1121-1126 . DOI: 10.1009-0959.2025.120014

Abstract

 In recent years the rapid development of new technologies particularly generative artificial intelligence has profoundly reshaped the foundational paradigms across various industries. In the specific field of drug regulation which is critical to public health and safety the introduction of AI technologies enhances the efficiency of review and approval routine oversight public service and decision-making support. However it also introduces unique ethical challenges due to its inherent complexity and unpredictability. This paper systematically outlines typical application scenarios of AI in drug regulation and provides an in-depth analysis of potential risks including inherent risks such as algorithmic bias and hallucinations), application risks such as model security vulnerabilities and data privacy breaches), and derivative risks such as blurred accountability and undermined regulatory credibility. By drawing on governance experiences of the World Health Organization UNESCO the European Union and other regulatory agencies and integrating with China's actual situation this study constructs a comprehensive ethical governance framework spanning four levels "concept-norm-governance-capability". This framework advocates for establishing a people-centered development-safety-balanced value orientation at the conceptual level promoting the integration of scientific and medical ethics requirements while developing specific technical standards at the normative level implementing a risk-based full-lifecycle regulatory mechanism to ensure traceability and accountability at the governance level and fostering a multidisciplinary talent pool and technical support platforms at the capability level. Through the establishment and refinement of this governance framework it aims to provide theoretical support and a practical framework for promoting the robust and responsible application of AI in the field of drug regulation in China thereby safeguarding public medication safety and health rights.

  

参考文献

    [1 高文.抢抓人工智能发展的历史性机遇——深刻领会习近平总书记关于人工智能的重要论述[J.中国信息化,2025,(2):6-9.

         2  张亚勤,霍福鹏.从“互联网+”到“人工智能+”迈向智能经济和智能社会发展新阶段[N.人民日报,2025-08-277.

         3  鲍华燕,罗杰.新时代药品监管落实“四个最严”要求几个重点问题探讨[J.中国药学杂志,2022575):408-412.

         4  国家新一代人工智能治理专业委员会.新一代人工智能伦理规范[EB/OL.2021-09-25)[2025-09-15.https//www.most.gov.cn/kjbgz/202109/t20210926_177063.htmlref=salesforce-research.

         5  王涛,宋海波,王青,等.WHO《卫生健康领域人工智能伦理与治理》指南简述与启示[J.中国药物警戒,2024218):906-909.

         6  国务院.国务院关于印发新一代人工智能发展规划的通知[EB/OL.2021-07-08)[2025-09-15.https//www.gov.cn/gongbao/content/2017/content_5216427.htm.

         7  外交部.全球人工智能治理倡议[EB/OL.2023-10-20)[2025-09-15.https//www.fmprc.gov.cn/web/ziliao_674904/1179_674909/202310/t20231020_11164831.shtml.

         8  2025世界人工智能大会暨人工智能全球治理高级别会议.人工智能全球治理行动计划[EB/OL.2025-07-26)[2025-09-15.https//www.gov.cn/yaowen/liebiao/202507/content_7033929.htm.

         9  国务院.国务院关于深入实施“人工智能+”行动的意见[EB/OL.2025-08-21)[2025-09-15.https//www.gov.cn/gongbao/2025/issue_12266/202509/content_7039598.html.

         10 中共中央办公厅,国务院办公厅.中共中央办公厅 国务院办公厅印发《关于加强科技伦理治理的意见》[EB/OL.2022-03-20)[2025-09-15.https//www.gov.cn/gongbao/content/2022/content_5683838.htm.

         11 科技部,教育部,工业和信息化部,等.关于印发《科技伦理审查办法(试行)》的通知[EB/OL.2023-09-07)[2025-09-15.https//www.most.gov.cn/xxgk/xinxifenlei/fdzdgknr/fgzc/gfxwj/gfxwj2023/202310/t20231008_188309.html.

         12 工业和信息化部.工业和信息化部等部门公开征求对《人工智能科技伦理管理服务办法(试行)》的意见[EB/OL.2025-08-07)[2025-09-15.https//www.miit.gov.cn/jgsj/kjs/jscx/gjsfz/art/2025/art_092a447008f340d3abd55819b8c8e5cf.html.

         13 科技部.发展负责任的人工智能:新一代人工智能治理原则发布[EB/OL.2019-06-17)[2025-09-15.https//www.most.gov.cn/kjbgz/201906/t20190617_147107.html.

         14 全国网络安全标准化技术委员会,国家计算机网络应急技术处理协调中心.人工智能安全治理框架2.0EB/OL.2025-09-15)[2025-09-15.https//www.cac.gov.cn/2025-09/15/c_1759653448369123.htm.

         15 国家药品监督管理局.药品监管人工智能典型应用场景清单[EB/OL.2024-06-18)[2025-09-15.https//www.nmpa.gov.cn/xxgk/fgwj/gzwj/gzwjzh/20240618144318144.html.

         16 张平.人工智能伦理治理研究[J.科技与法律(中英文),2024,(5):1-12.

         17 全国人民代表大会常务委员会.中华人民共和国药品管理法[EB/OL.2019-08-26)[2025-09-15.http//www.npc.gov.cn/npc/c2/c30834/201908/t20190826_300489.html.

         18 中国政府网.国家药监局国家卫生健康委关于发布药物临床试验质量管理规范的公告[EB/OL.2020-04-23)[2025-09-15.https//www.gov.cn/gongbao/content/2020/content_5525106.htm.

         19 世界医学会.世界医学会《赫尔辛基宣言》[EB/OL.2024-10-19)[2025-09-15.https//www.wma.net/wp-content/uploads/2024/12/DoH-2024-in-Simplified-Chinese.pdf.

         20 王玥,宋雅鑫,王艺霏,等.卫生领域人工智能的伦理与治理:多模态大模型指南[J.中国医学伦理学,2024379):1001-1022.

         21 世界卫生组织.世界卫生组织指南:卫生健康领域人工智能伦理与治理[EB/OL.2021)[2025-09-15.https//ai-ethics-and-governance.institute/wp-content/uploads/2022/04/WHO-Guidance-Ethics-Governance-AI-for-Health-Chinese-version.pdf.

         22 联合国教科文组织.人工智能伦理问题建议书[EB/OL.2021-11-23)[2025-09-15.https//unesdoc.unesco.org/ark/48223/pf0000381137_chi.

         23 NIST. Plan for Federal AI Standards EngagementEB/OL.2022-04-05)[2025-09-15.https//www.nist.gov/artificial-intelligence/plan-federal-ai- standards-engagement.

         24 木子婷.FDA整合AI监管资源[N.中国医药报, 2024-09-14.

         25 FDA. FDA Launches Agency-Wide AI Tool to Optimize Performance for the American PeopleEB/OL.2025-06-02)[2025-09-15.https//www.fda.gov/news-events/press-announcements/fda-launches-agency-wide-ai-tool-optimize-performance-american-people.

         26 European Union. Ethics Guidelines for Trustworthy Artificial IntelligenceEB/OL.2019-01-26)[2025-09-15.https//digitalstrategy.ec.europa.eu/en/library/ethics -guidelines-rustworthy-ai.

         27 European Union. The EU Artificial Intelligence ActEB/OL.2024-05-30)[2025-09-15.https//artificialintelligenceact.eu/.

         28 EMA/HMA.NDSG workplan 2025-2028 Data and AI in medicines regulation EB/OL.2025-05-07)[2025-09-15.https//www.ema.europa.eu/en/documents/other/network-data-steering-group-workplan-2025-2028_en.pdf.

文章导航

/