mlsanityData to 14 Oct 2026

Protecting patient privacy in clinical foundation models: Technical and legal perspectives

arXiv 2608.07705 ↗First posted 7 Aug 2026cs.AI cs.LGcited by 1

Not on the map. Why?

Abstract (arXiv)

Clinical foundation models trained on large-scale patient data are increasingly used for decision support, screening, and public health planning. As deployment expands, privacy risk arises from model-mediated leakage, yet its prevalence and severity remain poorly quantified. Models can disclose sensitive training artifacts, enabling patient re-identification in ways not captured by data-handling controls alone. As a result, existing frameworks, including HIPAA and GDPR, offer limited protection against assessing and addressing. We propose a practical framework for assessing privacy risk in clinical foundation models, illustrate realistic leakage scenarios across deployment settings, map them to legal regimes, and outline complementary technical and legal mitigations. Our analysis provides a context-aware risk assessment grounded in realistic usage to preserve the value of medical foundation models while rigorously safeguarding patient privacy.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 33 of its 78 references that are in the map. First posted on arXiv 7 Aug 2026.

Builds on

33 of 78 references.

References · 33

  1. 2016Deep Learning with Differential Privacycited by 2,728
  2. 2018The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networkscited by 864
  3. 2019Generalization through Memorization: Nearest Neighbor Language Modelscited by 800
  4. 2022Quantifying Memorization Across Neural Language Modelscited by 763
  5. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 446
  6. 2025MedGemma Technical Reportcited by 436
  7. 2023MedAlpaca -- An Open-Source Collection of Medical Conversational AI Models and Training Datacited by 286
  8. 2022GatorTron: A Large Clinical Language Model to Unlock Patient Information from Unstructured Electronic Health Recordscited by 238
  9. 2022Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Modelscited by 220
  10. 2024Safety Layers in Aligned Large Language Models: The Key to LLM Securitycited by 134
  11. 2023On Memorization in Diffusion Modelscited by 125
  12. 2020Synthetic Data -- Anonymisation Groundhog Daycited by 108
21 more
  1. 2024Foundation Model for Advancing Healthcare: Challenges, Opportunities, and Future Directionscited by 77
  2. 2024Zero Shot Health Trajectory Prediction Using Transformercited by 77
  3. 2025International AI Safety Reportcited by 67
  4. 2021Deep Learning-based Patient Re-identification Is able to Exploit the Biometric Nature of Medical Chest X-ray Datacited by 38
  5. 2019Disparate Vulnerability to Membership Inference Attackscited by 35
  6. 2021Source Inference Attacks in Federated Learningcited by 34
  7. 2023A Multi-Center Study on the Adaptability of a Shared Foundation Model for Electronic Health Recordscited by 31
  8. 2025Privacy Auditing of Large Language Modelscited by 23
  9. 2024Generative AI in EU Law: Liability, Privacy, Intellectual Property, and Cybersecuritycited by 21
  10. 2024Medical Multimodal Foundation Models in Clinical Diagnosis and Treatment: Applications, Challenges, and Future Directionscited by 18
  11. 2025The Canary's Echo: Auditing Privacy Risks of LLM-Generated Synthetic Textcited by 18
  12. 2025LLM Unlearning with LLM Beliefscited by 17
  13. 2025Hubble: a Model Suite to Advance the Study of LLM Memorizationcited by 17
  14. 2024De-identification is not enough: a comparison between de-identified and synthetic clinical notescited by 14
  15. 2025Pretraining with hierarchical memories: separating long-tail and common knowledgecited by 14
  16. 2026Large-scale online deanonymization with LLMscited by 14
  17. 2025Building Safe GenAI Applications: An End-to-End Overview of Red Teaming for Large Language Modelscited by 11
  18. 2025The Inadequacy of Offline LLM Evaluations: A Need to Account for Personalization in Model Behaviorcited by 8
  19. 2025Memorization Sinks: Isolating Memorization during LLM Trainingcited by 7
  20. 2022Current State of Community-Driven Radiological AI Deployment in Medical Imagingcited by 5
  21. 2025An Investigation of Memorization Risk in Healthcare Foundation Modelscited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Quantifying privacy risks in foundation model representations of brain MRIcited by 0