Public skepticism about the use of AI in scientific research

Generative AI tools are becoming common in scientific work, yet little is known about how the public – the taxpayers who fund much of this research – view their use. In a preregistered, nationally representative UK survey (N = 1,465), we find that the public is wary of AI-assisted research: While most respondents support public funding for scientific research in general (77%), support falls sharply when generative AI is involved (42%). This AI funding penalty is strongest among scientifically literate, politically liberal, younger, and female respondents. Moreover, 34–40% say they would trust both the science and the scientists less if generative AI was used during research. Although our findings do not indicate blanket rejection of AI in science, they point to a potential mismatch between scientific practice and public expectations, one that has implications for the social legitimacy of AI-assisted research and how it is governed and disclosed.