Attitudes and Perceptions of Generative Artificial Intelligence Chatbots in the Peer Review of Traditional, Complementary, and Integrative Medicine Research: A Large-Scale, International Cross-Sectional Survey
Year: 2026
Type: preprint
Abstract: Background Generative artificial intelligence (GenAI) chatbots are being explored for their potential to support the peer review process by improving efficiency and analytical capacity. While early applications show promise, concerns regarding ethics, privacy, and research integrity remain, and their use within traditional, complementary, and integrative medicine (TCIM) remains underexplored. This study examines TCIM researchers’ attitudes and perceptions regarding the use of GenAI chatbots in peer review through a large-scale international survey. Methods Corresponding authors of TCIM-related publications indexed in Ovid MEDLINE were invited to participate in an online, anonymous, international cross-sectional survey distributed via SurveyMonkey. The questionnaire collected data on demographic characteristics, the perceived importance of GenAI in peer review, the likelihood of adoption, and the perceived benefits and concerns. Results A total of 889 eligible respondents participated. Overall, 40.25% (n=285/708) reported a positive perception of GenAI in peer review, and 41.67% (n=295/708) considered it important to the field. 34.24% researchers reported being likely (n=253/739) or 18.27% (135/739) very likely (totaling 52.51%) to use GenAI in peer review, highlighting a gap between perceived importance and intended adoption. Identified benefits included improved efficiency (50.72%, n=350/690), reduced reviewer workload (51.16%, n=353/690), and generation of constructive feedback (46.30%, n=319/689), reflecting respondents selecting “agree.” Concerns included bias (49.85%, n=333/668), lack of transparency (42.86%, n=285/665), inaccuracies (49.40%, n=329/666), accountability challenges (45.20%, n=301/666), and data privacy risks (36.05%, n=239/663). Most respondents believed that peer reviewers require either some training (43.91%, n=310/706) or a lot of training (40.93%, n=289/706) to use GenAI chatbots effectively in peer review. Conclusion Respondents reported perceived benefits of GenAI in peer review, alongside notable concerns regarding its use. While opinions were mixed, findings suggest both perceived utility and apprehension about potential risks. Further work is needed to understand these perspectives better and to explore how GenAI can be responsibly integrated into peer-review processes.
Source: medRxiv
Authors: Jeremy Y. Ng, Jamie Tan, Karthik Adapa, Prashant Kumar Gupta, Madhur Gupta, Shao Li, Darshan H. Mehta, Melinda Ring, Manisha Shridhar, João Paulo Souza, Tetsuhiro Yoshino, Myeong Soo Lee, Holger Cramer
Institutions: University of Technology Sydney, Universitätsklinikum Tübingen, Bosch Health Campus, McMaster University, World Health Organization Regional Office for South-East Asia, All India Institute of Ayurveda, New Delhi, World Health Organization - India, Beijing National Research Center for Information Science and Technology, Tsinghua University, Brigham and Women's Hospital, Harvard University, Massachusetts General Hospital, Benson-Henry Institute, Northwestern University, World Health Organization Regional Office for the Americas, Pan American Health Organization (Brasil), Centro Latino-Americano e do Caribe de Informação em Ciências da Saúde, Keio University, Keio University Hospital, Korea Institute of Oriental Medicine
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Keywords: generative AI, AI chatbots, peer review, traditional, complementary and integrative medicine, researcher attitudes, technology adoption, research integrity, algorithmic bias, data privacy, reviewer training, AI in scientific publishing, research ethics, reviewer workload, transparency and accountability, AI hallucination
Open Access status: green
OpenAlex ID: W7106145251
