Article Text

Download PDFPDF
Survey of US physicians’ attitudes and knowledge of AI
  1. Sarah Gebauer1,
  2. Carly Eckert2
  1. 1Anesthesiology, Elk River Anesthesia Associates, Steamboat Springs, Colorado, USA
  2. 2University of Washington, Seattle, Washington, USA
  1. Correspondence to Dr Sarah Gebauer, Anesthesiology, Elk River Anesthesia Associates, Steamboat Springs, Colorado 80487, USA; sarah.l.gebauer{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Artificial intelligence (AI) applications in healthcare have proliferated over the past several years.1 Potential uses include decreased administrative burden, improved diagnostic accuracy and personalised treatment recommendations.2 Image-related fields such as ophthalmology and radiology have seen rapid uptake of AI tools, and the technology will likely continue to develop in other specialties.3 Clinically active physicians cannot keep up with the array of technological advances, their assumptions and implications. Understanding physician perspectives on AI tools and the level of confidence with these applications is critical as educational approaches are developed.

Physician sentiment about AI: early studies

Surveys of healthcare professionals have shown a combination of wariness and interest in AI tools. Early studies noted that distrust of AI among healthcare professionals may decrease its use and impede adoption, and that physicians doubt that AI will outperform humans in clinical tasks.4 One recent survey of 758 physicians and medical students showed support for using AI but concern about unpredictable results.5 In another survey of primary care healthcare professionals, half of whom were physicians, 66% reported no prior training in AI but 91% reported an interest.6

Several AI curricula for medical students and trainees have been published. Most focus on AI methodologies and refrain from topics such as bias and fairness.7 There is a lack of consensus about what elements are most …

View Full Text


  • SG and CE contributed equally.

  • Contributors Both authors actively collaborated on all parts of this paper including but not limited to survey development and deployment and paper development and editing.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.