As generative artificial intelligence (AI) tools become increasingly integrated into scientific research, questions arise about how such integration affects perceptions of legitimacy, accountability, and fairness in the production of scientific knowledge. This study investigates how the disclosure of AI use in academic research shapes public perceptions of researchers and their work. In a preregistered survey experiment (N = 806), respondents were randomly assigned to read one of six author disclosure statements describing different levels and types of generative AI use. Respondents reported the highest levels of trust, perceived research quality, and perceived ethicality when the research was described as conducted entirely without AI, and the lowest when AI was used for both theoretical and methodological tasks. Respondents were more positive toward the use of generative AI when it was used for methodological tasks such as coding compared with theoretical tasks such as hypothesis generation. “AI for language polishing” was perceived as least transparent. Frequency of AI use, but not self-reported AI knowledge or general trust in science, moderated these effects: individuals with more personal experience using AI were more accepting of AI-assisted research. These results indicated that public aversion to AI extended into the scientific domain and is driven less by ignorance than by concerns over diminished human creativity, effort, and accountability. Implications for maintaining trust and engagement in science amid increasing AI integration are discussed.
Chen et al. (Thu,) studied this question.