人工智慧（Artificial Intelligence, AI）技術近年快速發展，並廣泛應用於疾病篩檢、醫學影像判讀、臨床決策支援、風險預測及數位療法等醫療領域，使人工智慧醫療器材逐漸由傳統之輔助工具，發展為能夠參與醫療決策之重要技術。然而，人工智慧具有黑箱性及持續更新等技術特性，亦可能改變醫師與醫療機構之決策模式，進而對既有醫療過失責任法制帶來新的挑戰。
目前我國有關人工智慧醫療器材之研究，多著重於產品監管、資料治理及人工智慧技術發展等議題，而對於醫師及醫療機構使用人工智慧醫療器材所衍生之民事損害賠償責任，尚缺乏完整且具系統性之研究。本文爰以醫師及醫院使用人工智慧醫療器材之損害賠償責任為研究核心，首先介紹人工智慧醫療器材之定義、技術特性、演算模型及臨床應用，並探討人工智慧醫療器材之法律定性、醫療過失責任法理及醫療常規等基礎理論，進一步分析人工智慧導入醫療決策後，對醫師及醫療機構過失認定所產生之影響。
本文認為，相較於人工智慧之自主程度，以人工智慧於醫療決策中之人機互動模式作為責任分析架構，更能反映臨床實務之運作情形。因此，本文依人工智慧於臨床決策中之角色，區分為純輔助型、建議型與人機協作型，以及自主決策接近型三種態樣，並分別探討人工智慧黑箱性、資料偏移或模型偏移、醫師採納或偏離人工智慧建議，以及醫療機構管理義務等因素，對醫療過失判斷之影響。
本文研究結果認為，人工智慧醫療器材之導入並未改變我國醫療責任法之基本架構，醫療損害賠償責任仍應以醫師是否違反注意義務及醫療機構是否已盡管理義務作為核心判斷。然而，人工智慧之技術特性已使醫師注意義務之內容及醫療機構管理責任產生新的內涵。未來宜建立人工智慧醫療器材導入管理制度、人工智慧治理及使用紀錄機制，並持續完善臨床指引及醫療常規，以兼顧病人權益保障、醫療安全及人工智慧醫療創新之發展。

The rapid advancement of artificial intelligence (AI) has significantly transformed modern healthcare. AI-powered medical devices are now widely applied in disease screening, medical image interpretation, clinical decision support, risk prediction, and digital therapeutics, gradually evolving from conventional diagnostic aids into technologies capable of participating in clinical decision-making. Nevertheless, the technical characteristics of AI, including algorithmic opacity (the "black box" problem), model drift, data drift, and continuous learning, have introduced new challenges to the existing legal framework governing medical malpractice liability.
Current studies in Taiwan primarily focus on the regulation of AI medical devices, data governance, and technological development, while relatively limited attention has been paid to the civil liability arising from physicians' and hospitals' use of AI medical devices. Accordingly, this thesis examines the tort and contractual liability associated with the use of AI medical devices by physicians and healthcare institutions. It first reviews the definitions, technical characteristics, algorithmic models, and clinical applications of AI medical devices, followed by an analysis of their legal nature, the legal framework of medical malpractice, and the concept of the medical standard of care. Based on these foundations, this study explores how AI affects the determination of negligence in medical practice.
Rather than classifying AI according to its level of autonomy, this thesis argues that the degree of human-AI interaction more accurately reflects clinical practice and should serve as the analytical framework for allocating legal responsibility. Accordingly, AI medical devices are categorized into three types: assistive AI, advisory and collaborative AI, and near-autonomous AI. Within each category, this study examines how algorithmic opacity, data drift, model drift, physicians' acceptance or rejection of AI recommendations, and hospitals' management obligations may influence the assessment of medical negligence.
This thesis concludes that the introduction of AI medical devices does not fundamentally alter Taiwan's existing medical liability framework. The determination of liability should continue to focus on whether physicians have fulfilled their professional duty of care and whether healthcare institutions have properly discharged their management obligations. However, AI technologies have substantially reshaped the content of physicians' duties, the development of medical standards of care, and hospitals' governance responsibilities. To address these emerging challenges, Taiwan should establish comprehensive AI governance mechanisms, strengthen institutional management of AI medical devices, improve documentation and traceability systems, and continuously update clinical guidelines and standards of care. Such measures would facilitate the safe integration of AI into healthcare while balancing patient protection, medical safety, and technological innovation.

一、 中文（按作者姓氏筆畫排序）
(一) 書籍
1. 王澤鑑，侵權行為法，增補版，2023年。
2. 黃美鳳，醫事糾紛鑑定初鑑醫師指引手冊，2016年。
3. 陳春山，醫師、病人、醫療糾紛，三版，2004年。
(二) 翻譯書籍
1. Eric Topol（黃鈺閔、王欣薇、涂瑋瑛、李偉誠譯），DEEP MEDICINE 無情AI有情醫療，2020年。
(三) 專書論文
1. 吳振吉，人工智慧醫療應用之法律責任-以人工智慧之醫學影像判讀為切入點，收於：人工智慧與法律因應：以智慧醫療及智慧製造為中心，元照出版有限公司，頁85-112，2022年。
2. 吳振吉，醫療契約立法之比較研究（一）──典型契約化之考量因素及早期荷蘭立法之介紹〉，收於：實務家觀點下之醫療民事訴訟，頁33-54，2022年。
3. 游進發，醫療瑕疵判斷標準之研究，收於：醫師民事責任之實體與程序法上問題之研究：民事程序法焦點論壇第六卷，頁93-118，2019年。
(四) 期刊論文
1. 王一旅，深度學習AI 醫療民事責任之探討──以美國歐盟及我國法制為中心，輔仁法學，61期，頁1-85，2021年。
2. 王皇玉，醫療法第82條修正後醫療常規與合理臨床專業裁量之法界現況探討，臺灣醫學，28卷3期，頁330-336，2024。
3. 林宗穎，醫療機構組織責任之理論與案例類型之具體化──以德國與臺灣案例為中心，政大法學評論，148期，頁163-244，2017年。
4. 吳志正，變動中之民事醫療過失判定基準，政大法學評論，158期，頁87-159，2018年。
5. 吳志正，醫療契約之定性，月旦法學雜誌，139期，頁200-214，2006年。
6. 胡峰賓、翁紹仁，人工智慧醫療實務應用下之法制問題探討，全國律師，30：1期，頁2-14，2026年。
7. 胡凱君、顏孟鐸、王家鼎、徐士昌、許金旺，AI在醫學教育應用：有效運用人工智慧提升教與學，台灣急診醫學通訊，7卷5期，2024年。
8. 侯英泠，醫療機構之組織義務再建構，月旦民商法雜誌，74期，頁6-27，2021年。
9. 侯英泠，從SARS談醫院院內感控義務之契約責任問題，月旦法學雜誌，105期，頁48-66，2004年。
10. 陳聰富，醫療事故民事責任之過失判定，政大法學評論，127期，頁123-193，2011年。
11. 陳聰富，醫療法：第四講　醫療契約之法律關係（上），月旦法學教室，72期，頁87-98，2008年。
12. 陳聰富，醫療法：第四講　醫療契約之法律關係（下），月旦法學教室，73期，頁57-68，2008年。
13. 廖建瑜，醫療水準與醫療慣行之注意義務，月旦醫事法報告，10期，頁86-95，2017年。
14. 蕭輔仁、廖珩君、廖俊智，深度學習在腦部影像的臨床應用，臺灣醫學，24卷1期，頁50-57，2020年。
15. 盧映潔、葛建成、高忠漢，論醫療行爲之常規診療義務，臺大法學論叢，35卷4期，頁160-187，2026年。
(五) 研討會論文
1. 川口浩一，醫療AI與刑法上的行為概念-非人類之代理(agent)是否可以成為行為主體，第31次台北醫法論壇，財團法人人工智慧法律國際研究基金會主辦，頁53-57，2026年5月9日。
(六) 學位論文
1. 陳冠諭，醫療訴訟舉證責任分配與舉證減輕之研究──以醫療瑕疵、醫療過失與因果關係為中心，國立政治大學法律學系碩士論文，2023年1月。
(七) 網路資料
1. joycechin，北醫大醫療體系導入AI助攻護理交班，每月省下近千小時文書時間，《今日北醫》電子報，2026年4月15日，https://tmubt.tmu.edu.tw/2026/04/15/%E5%8C%97%E9%86%AB%E5%A4%A7%E9%86%AB%E7%99%82%E9%AB%94%E7%B3%BB%E5%B0%8E%E5%85%A5ai%E5%8A%A9%E6%94%BB%E8%AD%B7%E7%90%86%E4%BA%A4%E7%8F%AD%EF%BC%8C%E6%AF%8F%E6%9C%88%E7%9C%81%E4%B8%8B%E8%BF%91%E5%8D%83/。
2. 林裕洋，【中榮系列】以數據為本打造智慧醫療新典範，CIO Taiwan，2025年12月15日，https://www.cio.com.tw/103933/。
3. 林周義，15秒救一命！亞東醫院預警系統「AI醫療黑科技」躍登國際，臺灣智慧醫療創新整合平台，2026年3月11日，https://www.chinatimes.com/realtimenews/20260311001985-260418?chdtv。
4. 台北慈濟醫院，台北慈濟醫院參訪奇美醫院　觀摩AI智慧醫療應用與發展，2025年2月24日，https://taipei.tzuchi.com.tw/%E5%8F%B0%E5%8C%97%E6%85%88%E6%BF%9F%E9%86%AB%E9%99%A2%E5%8F%83%E8%A8%AA%E5%A5%87%E7%BE%8E%E9%86%AB%E9%99%A2-%E8%A7%80%E6%91%A9ai%E6%99%BA%E6%85%A7%E9%86%AB%E7%99%82%E6%87%89%E7%94%A8%E8%88%87/。
5. 奇美醫院智慧醫療中心，負責任 AI 執行中心，https://sub.chimei.org.tw/54020/index.php/ai/ai04/10-ai。
6. 奇美醫院智慧醫療中心，註137；亞東紀念醫院，人工智慧中心 部門目標，https://depart.femh.org.tw/medresearch/ai.html。
7. 亞東紀念醫院，人工智慧中心-部門目標，https://depart.femh.org.tw/medresearch/ai.html?utm。
8. 衛生福利部，醫療機構應用生成式人工智慧指引，2026年5月29日，https://www.mohw.gov.tw/cp-18-86695-1.html。
9. 衛生福利部食品藥物管理署，人工智慧/機器學習技術之醫療器材軟體查驗登記技術指引，2020年9月11日，https://www.fda.gov.tw/tc/siteListContent.aspx?sid=3787&id=34961。
10. 衛生福利部食品藥物管理署，醫用軟體分類分級指引，2022年9月15日，https://www.fda.gov.tw/tc/siteListContent.aspx?sid=310&id=41636。
11. 衛生福利部醫事司，114年度醫院評鑑基準(區域醫院、地區醫院適用)，2025年4月21日，https://dep.mohw.gov.tw/DOMA/cp-948-82244-106.html。
12. 游皓婷、侯力元，「電腦輸入錯」開錯刀醫師遭懲處，TVBS新聞網，https://news.tvbs.com.tw/life/96587。
13. 洪巧藍，全球最佳智慧醫院「台灣13家上榜」台中榮總連2年闖進百大，ETtoday健康雲，2025年9月12日，https://health.ettoday.net/news/3032285。
14. 臺北榮民總醫院醫療人工智慧發展中心，負責任AI執行中心，2026年5月20日，https://wd.vghtpe.gov.tw/maic/Fpage.action?muid=20314&fid=18349。
15. 臺北榮民總醫院，賀！本院榮獲27屆「國家生技醫療品質獎」二銀一銅九標章佳績，2025年1月24日，https://www.vghtpe.gov.tw/News%21one.action?nid=14231 。
16. 臺北醫學大學校級人工智慧醫療研究中心，中心介紹，https://aimc.tmu.edu.tw/ 。
17. 臺北醫學大學附設醫院，2019未來科技突破獎-遠距人工智慧重症照護平台，https://www.tmuh.org.tw/single/4960 。
18. 慧德科技，中榮人工智慧醫療軟體 Acura AKI 獲 TFDA 軟體醫療器材認證，2025年7月21日，https://www.huede.com.tw/portal_b1_page.php?button_num=b1&cnt_id=19292&owner_num=b1_80528&u。
19. 劉忠峰，「以人為本」的奇美智慧醫療，奇美醫訊，https://www.chimei.org.tw/main/cmh_department/59012/magazin/vol148/04-7.html。
20. 潘韜宇，大幅提升判讀率、準確度！亞東醫院智慧胸腔X光AI系統榮獲「國家醫療品質獎」金獎肯定，立報傳媒，2021年1月25日，https://www.limedia.tw/tech/17317/。
21. 廖家德，從Prompt到Promise：打造AI先行的智慧醫療，奇美醫訊，https://www.chimei.org.tw/main/cmh_department/59012/magazin/vol150/04-1.html。
22. 萬恩慈，2025台灣醫療科技展：AI應用與創新科技助攻臨床照護，亞東院訊，2026年1月，https://www.femh.org.tw/magazine/viewmag.aspx?ID=11870。
二、 英文
(一) 期刊論文
1. Bester JC, Cole CM & Kodish ED, The Limits of Informed Consent for an Overwhelmed Patient: Clinicians' Role in Protecting Patients and Preventing Overwhelm, 18 AMA J Ethics 869-886 (2016).
2. Bryson, J.J., Diamantis, M.E. & Grant, T.D., Of, for, and by the people: the legal lacuna of synthetic persons, 25 Artif Intell Law 280-282 (2017).
3. Gerke, S., Minssen, T. & Cohen, G., Ethical and Legal Challenges of Artificial Intelligence-Driven Healthcare, in Artificial Intelligence in Healthcare, 26 Artificial Intelligence in Healthcare 295-336 (2020).
4. Jiang, F. et al., Artificial Intelligence in Healthcare: Past, Present and Future, 2 Stroke Vasc Neurol. 230-243 (2017).
5. Jieun Oh, Artificial intelligence-driven real-time assistance in minimally invasive surgery: a technology-oriented narrative review, 29 J Minim Invasive Surg. 70-81 (2026).
6. Kreimeyer, K. et al., Natural Language Processing Systems for Capturing and Standardizing Unstructured Clinical Information: A Systematic Review, 73 Journal of Biomedical Informatics 14-29 (2017).
7. Little, S. et al., Adaptive Deep Brain Stimulation in Advanced Parkinson Disease, 74 Annals of Neurology 449-457 (2013).
8. McKinney S.M. et al., International evaluation of an AI system for breast cancer screening, 577 Nature 89-94.
9. Miner, A. S. et al., Smartphone-Based Conversational Agents and Responses to Questions About Mental Health, Interpersonal Violence, and Physical Health, 176 JAMA Intern Med. 619-625 (2016).
10. Phan.P et al., Digital therapeutics in the clinic, 8 Bioeng Transl Med 1-17(2023).
11. Perez, M. V. et al., Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation, 381 N Engl J 1909-1917 (2019).
12. Price, W. N. II, Gerke, S. & Cohen, I. G., Potential Liability for Physicians Using Artificial Intelligence, 322 JAMA 1765-1766 (2019).
13. Shanafelt, T. D. et al., Relationship Between Clerical Burden and Characteristics of the Electronic Environment with Physician Burnout and Professional Satisfaction, 91 Mayo Clin Proc. 836-848 (2016).
14. Solum, L. B., Legal Personhood for Artificial Intelligences, 70 North Carolina Law Review 1231-1287 (1992).
15. S G Fraser , W Adams, Wrong site surgery, 90 Br J Ophthalmol 814-816(2006).
(二) 網路資料
1. 3M Health Information Systems, Clinical Documentation Integrity (CDI), https://www.3m.com/3M/en_US/health-information-systems-us/solutions/clinical-documentation-integrity/.
2. Aidoc, Official Website, https://www.aidoc.com/platform.
3. Big Health, Sleepio, https://www.sleepio.com/.
4. BioBeat, Official Website, https://www.bio-beat.com/.
5. Cohen, I. G., Informed Consent and Medical Artificial Intelligence: What to Tell the Patient? (2020), https://www.law.georgetown.edu/georgetownlaw-journal/wp-content/uploads/sites/26/2020/06/Cohen_Informed-Consent-and-Medical-Artificial-Intelligence-What-to-Tell-the-Patient.pdf.
6. Digital Therapeutics Alliance (DTA), Digital Therapeutics Definition and Core Principles (2023), /DTA_DTx-Definition-and-Core-Principles.pdf.
7. European Commission, Ethics Guidelines for Trustworthy AI (2019), https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai.
8. European Parliament, Civil Liability Regime for Artificial Intelligence (2020), https://www.europarl.europa.eu/doceo/document/TA-9-2020-0276_EN.html.
9. German Federal Institute for Drugs and Medical Devices (BfArM), DiGA Directory, https://diga.bfarm.de/en.
10. International Medical Device Regulators Forum (IMDRF), Software as a Medical Device (SaMD): Key Definitions (2013), https://www.imdrf.org/documents/software-medical-device-samd-key-definitions.
11. International Medical Device Regulators Forum (IMDRF), Software as a Medical Device (SaMD): Possible Framework for Risk Categorization and Corresponding Considerations (2014), https://www.imdrf.org/documents/software-medical-device-samd-possible-framework-risk-categorization-and-corresponding.
12. International Organization for Standardization (ISO), What is artificial intelligence (AI)?, https://www.iso.org/artificial-intelligence/what-is-ai.
13. Intuitive Surgical, da Vinci Surgical System, https://www.intuitive.com/en-us/products-and-services/da-vinci.
14. Joint Commission International (JCI), JCI Accreditation Standards for Hospitals (8th ed., 2024), https://www.jointcommissioninternational.org/standards/.
15. Medicines and Healthcare products Regulatory Agency (MHRA), Software and AI as a Medical Device Change Programme (2021/09/16), https://www.gov.uk/government/publications/software-and-ai-as-a-medical-device-change-programme.
16. Medicines and Healthcare products Regulatory Agency (MHRA), AI Airlock Sandbox Pilot Programme Report(2025/11/25), https://www.gov.uk/government/publications/ai-airlock-sandbox-pilot-programme-report.
17. Medtronic, Deep Brain Stimulation Therapy, https://www.medtronic.com/en-us/patients/treatments-therapies/deep-brain-stimulation.html.
18. Nuance Communications, DAX Copilot, https://www.nuance.com/healthcare/ambient-clinical-intelligence/dax-copilot.html.
19. Paige, Official Website, https://paige.ai/.
20. Pear Therapeutics, Official Website, https://peartherapeutics.com/.
21. Price, W. N. II, Medical Malpractice and Black-Box Medicine (2018), https://repository.law.umich.edu/book_chapters/169/.
22. Qventus, Solutions, https://www.qventus.com/solutions/.
23. Sullivan, H. R. & Schweikart, S. J., Are Current Tort Liability Doctrines Adequate for Addressing Injury Caused by AI? (2019), https://journalofethics.ama-assn.org/article/are-current-tort-liability-doctrines-adequate-addressing-injury-caused-ai/2019-02.
24. The Organisation for Economic Co-operation and Development (OECD), OECD Recommendation of the Council on Artificial Intelligence (2019), https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449.
25. U.S. Food and Drug Administration (FDA), 21st Century Cures Act (2016), https://www.fda.gov/regulatory-information/selected-amendments-fdc-act/21st-century-cures-act.
26. U.S. Food and Drug Administration (FDA), Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD) Action Plan (2021), https://www.fda.gov/media/145022/download.
27. U.S. Food and Drug Administration (FDA), Clinical Decision Support Software: Guidance for Industry and Food and Drug Administration Staff (2024), https://www.fda.gov/regulatory-information/search-fda-guidance-documents/clinical-decision-support-software.
28. U.S. Food and Drug Administration (FDA), Device Software Functions Including Mobile Medical Applications: Guidance for Industry and Food and Drug Administration Staff (2022), https://www.fda.gov/regulatory-information/search-fda-guidance-documents/device-software-functions-including-mobile-medical-applications.
29. U.S. Food and Drug Administration (FDA), Marketing Submission Recommendations for a Predetermined Change Control Plan for AI-Enabled Device Software Functions (2025), https://www.fda.gov/regulatory-information/search-fda-guidance-documents/marketing-submission-recommendations-predetermined-change-control-plan-artificial-intelligence.
30. U.S. Food and Drug Administration (FDA), Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD): Discussion Paper and Request for Feedback (2019), https://www.fda.gov/media/122535/download.
31. Woebot Health, Woebot Health Platform, https://woebothealth.com/platform/.
32. World Health Organization (WHO), Ethics and Governance of Artificial Intelligence for Health (2021), https://www.who.int/publications/i/item/9789240029200.
33. World Health Organization (WHO), Strengthening Diagnosis and Treatment of Diabetic Retinopathy in the South-East Asia Region (2020), https://www.who.int/publications/i/item/9789290227946.
34. World Health Organization (WHO), WHO Surgical Safety Checklist (First Edition), https://www.who.int/docs/default-source/patient-safety/9789241598590-eng-checklist.pdf.