Background: Artificial intelligence (AI) tools have rapidly entered academic and clinical workflows, offering new ways to streamline documentation, enhance writing clarity, and reduce administrative burden. Within medical education, AI is increasingly used for tasks such as editing, summarizing, and generating written content. One emerging area is the use of AI to assist in drafting or refining letters of recommendation (LORs), a core component of residency and fellowship applications. Despite its growing prevalence, little is known about how hospitalist faculty perceive AI use in this context, how frequently they incorporate AI tools into LOR writing, or what benefits and risks they associate with this practice. Clarifying these perceptions is important as institutions consider developing guidance to support ethical, consistent, and appropriate use of AI in academic recommendation processes. This study aimed to characterize hospitalist attitudes, utilization patterns, and concerns regarding AI-assisted LOR writing.
Methods: A cross-sectional survey was distributed to 47 faculty hospitalists at a single academic institution. The survey assessed frequency and type of AI use within the past two years, perceived benefits and concerns, comfort with AI-assisted writing tasks, and attitudes toward the need for institutional guidelines. Descriptive statistics were used to summarize responses.
Results: Among respondents, 72% found AI acceptable for editing or refining written content, and 57% reported using AI for grammar and clarity improvements. Despite this, only 36% had used AI specifically for LOR writing, with ChatGPT identified as the most commonly used platform. Reported benefits of AI-assisted LOR writing included time savings (77%), assistance with overcoming writer’s block (57%), improved clarity (51%), and enhanced organization (45%).Faculty also expressed substantial concerns: loss of personal authenticity (83%) and ethical or professionalism considerations (45%) were the most frequently cited. A majority (74%) supported the development of institutional guidelines to ensure appropriate and ethical use of AI in LOR writing.
Conclusions: While faculty recognize the value of AI in improving efficiency, clarity, and workflow during LOR writing, concerns remain regarding authenticity, professionalism, and responsible use. Most respondents support the development of institutional guidance to ensure ethical integration of AI tools. As AI becomes more prevalent in academic writing, further research is needed to evaluate its long-term effects on the quality, credibility, and perception of letters of recommendation. Our next steps include developing institutional guidelines for appropriate utilization of AI in LOR writing as a pilot toward a more standardized process.