Abstract
Background
How clinicians conceptualize artificial intelligence reveals underlying assumptions about professional authority and decision-making. This study examined whether surgical and medical specialties frame AI differently in research and if such differences reflect divergent professional norms.
Methods
1561 AI-related research abstracts published between January 1, 2019, and March 27, 2025, in 30 high-impact journals. Abstracts were identified through a structured PubMed query and analyzed using a large language model (DeepSeek Reasoner) trained to classify along three dimensions: the human-AI relationship, the impact on professional autonomy, and the locus of decision control. A stratified validation sample was independently coded by a human rater. Chi-square testing and logistic regression were used to assess differences by specialty and publication year.
Results
Surgical abstracts more frequently framed AI as assistive (69.8% vs 54.9%; P < .001), explicitly addressed professional autonomy (73.5% vs 61.3%; P < .001), and specified decision control (69.3% vs 58.6%; P < .001) compared to medical abstracts. These differences persisted across the 7-year period. In multivariable logistic regression, assistive framing (OR, 2.43; 95% CI, 1.82-3.23) and explicit autonomy discussion (OR, 1.46; 95% CI, 1.11-1.92) were independently associated with surgical specialty.
Conclusions
Surgical and medical specialties exhibit distinct patterns in how they conceptualize AI, reflecting established perspectives on authority, expertise, and the human-machine relationship. These framings have implications for AI tool design, clinical implementation, and healthcare governance. Recognizing conceptual differences on AI is critical as healthcare transitions toward algorithmically mediated decision-making, as they may shape the future culture of clinical care.
Get full access to this article
View all access options for this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
