{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T14:49:02Z","timestamp":1785422942055,"version":"3.56.0"},"reference-count":35,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T00:00:00Z","timestamp":1782777600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Digit. Health"],"abstract":"<jats:p>The integration of Large Language Models (LLMs) into clinical workflows has blurred the boundary between passive decision-support tools and systems capable of autonomous diagnostic reasoning. Current regulatory instruments\u2014including the US FDA Clinical Decision Support Software (CDSS) guidance and the EU Artificial Intelligence Act (Regulation EU 2024\/1689)\u2014classify these systems primarily as \u201cdecision support,\u201d yet the non-deterministic outputs and emergent reasoning of generative AI introduce liability complexities these frameworks were not designed to address. This hypothesis and theory article examines the regulatory gap surrounding LLM-associated clinical errors. Drawing on regulatory precedents, case law, and a representative scenario of polypharmacy mismanagement, we analyze the risks of automation bias and the structural inadequacy of the Learned Intermediary Doctrine when applied to generative AI. We advance the autonomy\u2013liability correspondence hypothesis: legal liability should be allocated as a monotonic function of a system's structural autonomy from meaningful human review. From this we derive a \u201cThree-Tiered Liability Escalation Framework\u201d (T-LEF) that allocates accountability among AI developers, healthcare institutions, and clinicians according to measurable degrees of autonomy. Arguing that traditional Explainable AI is technically insufficient for foundation models, we propose Clinical Algorithmic Audit Trails (CAAT)\u2014a cryptographically secured, privacy-preserving traceability infrastructure\u2014as a complement to explainability, uncertainty estimation, human-factors evaluation, and post-market surveillance, alongside a structured Safe Harbor provision to support compliant innovation. Because paediatric patients\u2014from neonates to adolescents\u2014constitute a population in which these failure modes are both more probable and more consequential, we give paediatric care explicit focus: under-representation of paediatric data in training corpora, weight- and development-based dosing across narrow therapeutic windows, and children's inability to detect or contest erroneous recommendations jointly intensify these concerns, and a dedicated section examines how the T-LEF, the Clinically Deceptive Hallucination (CDH) evidentiary test, and CAAT apply to paediatric deployment. We acknowledge that the T-LEF and CAAT remain prospective proposals requiring empirical validation, stakeholder consultation, and refinement before adoption. This framework is offered as a structured contribution to regulatory deliberation, not a finalized implementation blueprint.<\/jats:p>","DOI":"10.3389\/fdgth.2026.1849190","type":"journal-article","created":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:38:22Z","timestamp":1782797902000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["From \u201cassistant\u201d to \u201cautonomous\u201d: legal liability and ethical traceability frameworks for generative AI in clinical misdiagnosis scenarios, with a special focus on paediatrics"],"prefix":"10.3389","volume":"8","author":[{"given":"Shiyi","family":"Xu","sequence":"first","affiliation":[{"name":"Capital Medical University Second Clinical School, Capital Medical University","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,6,30]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1038\/s41586-023-06291-2","article-title":"Large language models encode clinical knowledge","volume":"620","author":"Singhal","year":"2023","journal-title":"Nature"},{"key":"B2","doi-asserted-by":"publisher","first-page":"1930","DOI":"10.1038\/s41591-023-02448-8","article-title":"Large language models in medicine","volume":"29","author":"Thirunavukarasu","year":"2023","journal-title":"Nat Med"},{"key":"B3","doi-asserted-by":"publisher","first-page":"121","DOI":"10.1136\/amiajnl-2011-000089","article-title":"Automation bias: a systematic review of frequency, effect mediators, and mitigators","volume":"19","author":"Goddard","year":"2012","journal-title":"J Am Med Inform Assoc"},{"key":"B4","doi-asserted-by":"publisher","first-page":"e231938","DOI":"10.1001\/jamahealthforum.2023.1938","article-title":"ChatGPT and physicians' Malpractice risk","volume":"4","author":"Mello","year":"2023","journal-title":"JAMA Health Forum"},{"key":"B5","doi-asserted-by":"publisher","first-page":"1351","DOI":"10.1001\/jama.2023.5321","article-title":"AI-Generated medical advice\u2014GPT and beyond","volume":"329","author":"Haupt","year":"2023","journal-title":"JAMA"},{"key":"B6","volume-title":"Manufacturer and User Facility Device Experience (MAUDE) Database","year":"2024"},{"key":"B7","volume-title":"Clinical Decision Support Software: Guidance for Industry and Food and Drug Administration Staff","year":"2026"},{"key":"B8","volume-title":"Artificial Intelligence and Machine Learning (AI\/ML)-Based Software as a Medical Device (SaMD) Action Plan","year":"2021"},{"key":"B9","article-title":"Regulation (EU) 2024\/1689 laying down harmonised rules on artificial intelligence (artificial intelligence act)","year":"2024","journal-title":"OJEU"},{"key":"B10","article-title":"Directive (EU) 2024\/2853 of the European parliament and of the council of 23 October 2024 on liability for defective products and repealing council directive 85\/374\/EEC","year":"2024","journal-title":"OJEU"},{"key":"B11","volume-title":"Commission Work Programme 2025: Annex IV \u2014 intended Withdrawals. COM(2025) 45 Final","year":"2025"},{"key":"B12","doi-asserted-by":"publisher","first-page":"e195","DOI":"10.1016\/S2589-7500(20)30292-2","article-title":"Approval of artificial intelligence and machine learning-based medical devices in the USA and Europe (2015\u20132021): a comparative analysis","volume":"3","author":"Muehlematter","year":"2021","journal-title":"Lancet Digit Health"},{"key":"B13","doi-asserted-by":"publisher","first-page":"1765","DOI":"10.1001\/jama.2019.15064","article-title":"Potential liability for physicians using artificial intelligence","volume":"322","author":"Price","year":"2019","journal-title":"JAMA"},{"key":"B14","author":"Duffourc","year":""},{"key":"B15","doi-asserted-by":"publisher","first-page":"E160","DOI":"10.1001\/amajethics.2019.160","article-title":"Are current tort liability doctrines adequate for addressing injury caused by AI?","volume":"21","author":"Sullivan","year":"2019","journal-title":"AMA J Ethics"},{"key":"B16","doi-asserted-by":"publisher","first-page":"629","DOI":"10.1111\/1468-0009.12504","article-title":"Artificial intelligence and liability in medicine: balancing safety and innovation","volume":"99","author":"Maliha","year":"2021","journal-title":"Milbank Q"},{"key":"B17","year":""},{"key":"B18","doi-asserted-by":"publisher","first-page":"e11510","DOI":"10.2196\/11510","article-title":"Patient and consumer safety risks when using conversational assistants for medical information","volume":"20","author":"Bickmore","year":"2018","journal-title":"J Med Internet Res"},{"key":"B19","year":""},{"key":"B20","doi-asserted-by":"publisher","first-page":"423","DOI":"10.1093\/jamia\/ocw105","article-title":"Automation bias and verification complexity: a systematic review","volume":"24","author":"Lyell","year":"2017","journal-title":"J Am Med Inform Assoc"},{"key":"B21","doi-asserted-by":"publisher","first-page":"93","DOI":"10.1016\/j.ahj.2021.12.012","article-title":"Clinical events classification (CEC) in clinical trials: report on the current landscape and future directions \u2014 proceedings from the CEC summit 2018","volume":"246","author":"Sharma","year":"2022","journal-title":"Am Heart J"},{"key":"B22","doi-asserted-by":"publisher","first-page":"120","DOI":"10.1038\/s41746-023-00873-0","article-title":"The imperative for regulatory oversight of large language models (or generative AI) in healthcare","volume":"6","author":"Mesk\u00f3","year":"2023","journal-title":"npj Digit Med"},{"key":"B23","volume-title":"Restatement (Third) of Torts: Products Liability","year":"1998"},{"key":"B24","volume-title":"Restatement (Third) of Torts: Apportionment of Liability","year":"2000"},{"key":"B25","year":""},{"key":"B26","volume-title":"Quality Management System Regulation (QMSR), 21 CFR Part 820","year":"2026"},{"key":"B27","doi-asserted-by":"publisher","first-page":"310","DOI":"10.1186\/s12911-020-01332-6","article-title":"Explainability for artificial intelligence in healthcare: a multidisciplinary perspective","volume":"20","author":"Amann","year":"2020","journal-title":"BMC Med Inform Decis Mak"},{"key":"B28","doi-asserted-by":"publisher","first-page":"e745","DOI":"10.1016\/S2589-7500(21)00208-9","article-title":"The false hope of current approaches to explainable artificial intelligence in health care","volume":"3","author":"Ghassemi","year":"2021","journal-title":"Lancet Digit Health"},{"key":"B29","doi-asserted-by":"publisher","first-page":"e34407","DOI":"10.1016\/j.heliyon.2024.e34407","article-title":"Blockchain-enabled EHR access auditing: enhancing healthcare data security","volume":"10","author":"Ullah","year":"2024","journal-title":"Heliyon"},{"key":"B30","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1561\/0400000042","article-title":"The algorithmic foundations of differential privacy","volume":"9","author":"Dwork","year":"2014","journal-title":"Found Trends Theor Comput Sci"},{"key":"B31","volume-title":"Ethics and Governance of Artificial Intelligence for Health: Guidance on Large Multi-modal models","year":"2024"},{"key":"B32","doi-asserted-by":"publisher","first-page":"239","DOI":"10.1016\/S1473-3099(13)70250-0","article-title":"Diagnostic point-of-care tests in resource-limited settings","volume":"14","author":"Drain","year":"2014","journal-title":"Lancet Infect Dis"},{"key":"B33","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-57959-7","volume-title":"The EU General Data Protection Regulation (GDPR): A Practical Guide","author":"Voigt","year":"2017"},{"key":"B34","doi-asserted-by":"publisher","first-page":"313","DOI":"10.1001\/jamapediatrics.2023.5750","article-title":"Diagnostic accuracy of a large language model in pediatric case studies","volume":"178","author":"Barile","year":"2024","journal-title":"JAMA Pediatr"},{"key":"B35","doi-asserted-by":"publisher","first-page":"2114","DOI":"10.1001\/jama.285.16.2114","article-title":"Medication errors and adverse drug events in pediatric inpatients","volume":"285","author":"Kaushal","year":"2001","journal-title":"JAMA"}],"container-title":["Frontiers in Digital Health"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fdgth.2026.1849190\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:38:24Z","timestamp":1782797904000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fdgth.2026.1849190\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,30]]},"references-count":35,"alternative-id":["10.3389\/fdgth.2026.1849190"],"URL":"https:\/\/doi.org\/10.3389\/fdgth.2026.1849190","relation":{},"ISSN":["2673-253X"],"issn-type":[{"value":"2673-253X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,30]]},"article-number":"1849190"}}