Doctoral researchers’ Generative AI use and research-integrity concerns: a cross-context study

Generative artificial intelligence (GenAI) is embedded in doctoral research, raising concerns beyond general technology anxiety or coursework-focused academic integrity. This study examines doctoral researchers’ AI-related research-integrity concerns in relation to comfort with high-stakes AI use, perceived need for additional institutional AI guidance, and current study context. Data from Nature’s 2025 Graduate Survey were analysed for 2,099 doctoral researchers in science across selected East Asian and Western contexts. OLS models with robust standard errors were estimated for both outcomes, followed by interaction and robustness analyses. High-stakes AI comfort was associated with lower concern levels of both concern outcomes, whereas perceived need for additional institutional AI guidance was associated with higher levels of both outcomes. The negative comfort–concern association was weaker and non-significant in selected East Asian contexts. The findings highlight the importance of interpreting doctoral researchers’ AI-related research-integrity concerns in relation to doctoral-study context and perceived guidance needs.