EUVIMEDOpdag EUVIMED
Open access

Shadow AI and the Limits of Institutional Agility

Benjamin Rader, John S Brownstein

Journal of Medical Internet Research · 2026-10-14 · 28 · e112827-e112827

DOI: 10.2196/112827 ↗

At a glance

Journal article in Journal of Medical Internet Research, published online on 2026-10-14. Written by Benjamin Rader and 1 co-authors. Institution: Innovation and Digital Health Accelerator, Boston Children's Hospital, 300 Longwood Ave, Boston, MA, 02115, United States, 1 6173551389. Major MEDLINE topics: Artificial Intelligence. 10 references. The full text is freely available in PubMed Central.

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Physicians are adopting generative AI tools faster than health systems can evaluate them. This commentary on the empirical report of “Shadow AI” drivers by Petersson and colleagues argues that governing Shadow AI requires health systems to pair institutional speed with transparent, stakeholder-engaged governance.

Subjects

MeSH terms (MEDLINE)

Artificial IntelligenceGenerative Artificial IntelligenceHumans

Author keywords

clinical decision supportclinician workaroundsgenerative artificial intelligencehealth system governancelarge language modelsmedical device regulationregulatory sandbox

Authors and institutions

  • Innovation and Digital Health Accelerator, Boston Children's Hospital, 300 Longwood Ave, Boston, MA, 02115, United States, 1 6173551389.
  • Harvard Medical School, Boston, MA, United States.

Cite

APA

Rader, B., & Brownstein, J. S. (2026). Shadow AI and the Limits of Institutional Agility. Journal of Medical Internet Research, 28, e112827-e112827. https://doi.org/10.2196/112827

Vancouver

Rader B, Brownstein JS. Shadow AI and the Limits of Institutional Agility. Journal of Medical Internet Research. 2026;28:e112827-e112827. doi:10.2196/112827

BibTeX
@article{Rader2026,
  author = {Rader, Benjamin and Brownstein, John S},
  title = {Shadow AI and the Limits of Institutional Agility},
  journal = {Journal of Medical Internet Research},
  year = {2026},
  volume = {28},
  pages = {e112827--e112827},
  doi = {10.2196/112827}
}
Start for free →