Cover Image
QR Code

Federation Opacity as Governance Opportunity: toward a Fiduciary Model of Distributed Medical AI.

The article focuses on the concept of "federation opacity" in federated learning (FL) networks, particularly in healthcare, where the full training dataset remains inaccessible due to distributed data ownership. It argues that federation opacity is not merely an epistemic issue but a jurisdictional...

Full description

Published in:American Journal of Bioethics Vol. 26; no. 8; pp. 147 - 151
Main Author: Yang, Y. Tony
Format: commentary Journal Article
Published: Taylor & Francis Ltd Aug2026
Online Access:View this record in EBSCOhost
fields @attributes:
  recordID: 1
pdfLink:
plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=195895380&site=ehost-live
header:
  @attributes:
    shortDbName: ccm
    uiTerm: 195895380
    longDbName: CINAHL Complete
    uiTag: AN
  controlInfo:
    bkinfo:
    dissinfo:
    jinfo:
      jid:
        15265161
        FKZ
      jtl: American Journal of Bioethics
      issn: 15265161
      maglogo: N
    pubinfo:
      dt: Aug2026
      vid: 26
      iid: 8
      pid: 377
      pub: Taylor & Francis Ltd
      place: Philadelphia, Pennsylvania
    artinfo:
      ui:
        195895380
        195895380
        195895380
        10.1080/15265161.2026.2690932
        195895380
      ppf: 147
      ppct: 4
      formats:
        fmt:
          – @attributes:
              type: T
          – @attributes:
              type: P
      tig:
        atl: Federation Opacity as Governance Opportunity: toward a Fiduciary Model of Distributed Medical AI.
      aug:
        au: Yang, Y. Tony
        affil: The George Washington University
      sug:
        subj:
          Federated Learning Ethical Issues
          Data Security
          Hospitals
          Privacy and Confidentiality
          Information Storage
          Information Retrieval
          Systems Design
          Machine Learning Algorithms
      ab: The article focuses on the concept of "federation opacity" in federated learning (FL) networks, particularly in healthcare, where the full training dataset remains inaccessible due to distributed data ownership. It argues that federation opacity is not merely an epistemic issue but a jurisdictional challenge arising from divided institutional authority, which cannot be resolved solely through transparency or explainable AI. The author proposes establishing an independent Federated Learning Trust (FLT), a nonprofit fiduciary body with authority to represent patient-contributors, audit federations under controlled conditions, and certify ethically permissible FL use cases. This governance model aims to balance privacy preservation with accountability, address consent complexities, cybersecurity risks, regulatory disparities, and fairness concerns, thereby enhancing the legitimacy and ethical oversight of FL systems in medical contexts.
      pubtype: Academic Journal
      doctype:
        commentary
        Journal Article
      ougenre: Article
    language: English
    refInfo:
    holdings:
      @attributes:
        islocal: N