CHATGPT IN ACADEMIC RESEARCH: ADOPTION PATTERNS, ATTITUDES, AND ETHICAL PERCEPTIONS AMONG PAKISTANI RESEARCHERS

Abstract

The rapid emergence of generative artificial intelligence (AI), particularly ChatGPT, has transformed academic workflows and raised critical questions regarding its role in scholarly research. Despite growing global discourse, empirical evidence on researchers’ adoption patterns and ethical perceptions remains limited, especially in low- and middle-income countries. This study aimed to assess researchers’ awareness, adoption, attitudes, and ethical perceptions regarding ChatGPT in academic research, and to identify key predictors of its use. A national cross-sectional web-based survey was conducted among 237 researchers affiliated with universities and research institutions across Pakistan. A structured questionnaire evaluated awareness, usage patterns, perceived benefits, and ethical concerns related to ChatGPT. Descriptive statistics and multivariable logistic regression analyses were performed to examine factors associated with ChatGPT adoption. Overall, 67.1% of participants were aware of ChatGPT, while only 11.4% reported using it in academic research. Use was primarily limited to low-risk tasks such as paraphrasing and literature searching. Approximately half of respondents perceived ChatGPT as useful for improving research efficiency, whereas 46.4% expressed ethical concerns related to accountability and research integrity. Notably, 37.1% considered that ChatGPT could qualify for authorship under certain conditions. Multivariable analysis indicated that prior familiarity with chatbot technologies was a strong independent predictor of use (adjusted odds ratio [AOR] = 15.62, 95% CI 7.45–32.78, p < .001), while younger age was modestly associated with higher adoption (AOR = 1.08 per year decrease, 95% CI 1.01–1.16, p = .03). Despite widespread awareness, the adoption of ChatGPT in academic research remains limited and cautious. Ethical uncertainty, concerns regarding reliability, and lack of clear institutional guidance appear to constrain its integration. These findings highlight the need for structured AI literacy training, explicit disclosure policies, and robust governance frameworks to ensure responsible and transparent use of generative AI in scholarly research, particularly in resource-constrained settings.