Skip to content
DraconiticResearch
Search Sources API
186,541 studies31 sourcesUpdated 58 min ago
Peer-reviewedCC BY-NC-NDnpj Digital Medicine

Meaningful oversight of medical AI beyond human in the loop

  • Davy van de Sande1,2
  • Nicoleta Economou-Zavlanos3
  • Michel E. van Genderen4,5
5 affiliations
  1. 1Department of Adult Intensive Care, Erasmus MC University Medical Center, Rotterdam, The Netherlands
  2. 2Erasmus MC Datahub, University Center Rotterdam, Rotterdam, The Netherlands
  3. 3Duke University School of Medicine, Durham, NC, USA
  4. 4Department of Adult Intensive Care, Erasmus MC University Medical Center, Rotterdam, The Netherlands. [email protected]
  5. 5Department of Biostatistics and Bioinformatics, Duke University School of Medicine, Durham, NC, USA. [email protected]
Open at sourcePDFDOICite

Abstract

Human oversight of medical AI is increasingly required, but clinician presence alone does not make oversight meaningful. We propose four interlocking conditions—epistemic capacity, cognitive space, decisional authority, and intervention effectiveness—that determine whether human judgment can function as a safety mechanism. Embedding these conditions across procurement, deployment, monitoring, and decommissioning can turn oversight into an operational patient-safety capability across predictive, generative, semi-autonomous, and agentic systems.

Related

  1. Balancing Timely Access and Human Oversight in Patient-Facing Artificial Intelligence—Waiting for the Human in the Loop
    Felix Busch, Dennis M. Hedderich
    JAMA Network OpenPeer-reviewedCC BY 4.0Epidemiology & public healthCrossref
  2. Human-in-the-loop validation of a sequential multi-LLM medical education pipeline
    Yoojin Nam, Taein An, Sung Il Hwang +18
    NPJ Digital MedicinePeer-reviewedCC BY-NC-ND 4.0Full text1 citedEpidemiology & public healthPMC
  3. Public Perceptions of Timeliness for AI Use and Human Oversight.
    Natalie C Benda, Uday Suresh, Afra Shamnath +5
    JAMA network openPeer-reviewedCC BY-NC-NDFull text1 citedEpidemiology & public healthEurope PMCCrossref
  4. Beyond the Rankin Scale: A Patient‐Centered Score for Meaningful Recovery After Mechanical Thrombectomy in Older Adults With Pre‐Stroke Disability
    Hiroyasu Inoue, Masahiro Oomura, Yoshitaka Nakamura +16
    Geriatrics & gerontology internationalPeer-reviewedCC BYFull textGeneral medicineEurope PMC
  5. From clinician review to measurable human oversight of AI-drafted paediatric discharge documents
    D. Morquin, G. Bochu, L. Robert +7
    medRxivPreprintCC BYFull textEpidemiology & public healthmedRxiv

Cite

BibTeXRIS
BibTeX
@article{sande2026meaningful,
  title = {{Meaningful oversight of medical AI beyond human in the loop}},
  author = {Davy van de Sande and Nicoleta Economou-Zavlanos and Michel E. van Genderen},
  journal = {npj Digital Medicine},
  year = {2026},
  month = jul,
  doi = {10.1038/s41746-026-02971-1},
  url = {https://doi.org/10.1038/s41746-026-02971-1},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Meaningful oversight of medical AI beyond human in the loop
AU  - Davy van de Sande
AU  - Nicoleta Economou-Zavlanos
AU  - Michel E. van Genderen
PY  - 2026
DA  - 2026/07/23
JO  - npj Digital Medicine
DO  - 10.1038/s41746-026-02971-1
UR  - https://doi.org/10.1038/s41746-026-02971-1
AB  - Human oversight of medical AI is increasingly required, but clinician presence alone does not make oversight meaningful. We propose four interlocking conditions—epistemic capacity, cognitive space, decisional authority, and intervention effectiveness—that determine whether human judgment can function as a safety mechanism. Embedding these conditions across procurement, deployment, monitoring, and decommissioning can turn oversight into an operational patient-safety capability across predictive, generative, semi-autonomous, and agentic systems.
SN  - 2398-6352
ER  -