Public Perceives AI Scientists As Less Prudent Than Climate Scientists
The national survey explored how AI compares to other scientific fields across key indicators of trustworthiness.
Complete the form below to unlock access to ALL audio articles.
A new study from the Annenberg Public Policy Center (APPC) at the University of Pennsylvania has found that people in the United States view artificial intelligence (AI) science and scientists more negatively than climate scientists or scientists in general. The findings are based on a national survey exploring how AI compares to other scientific fields across key indicators of trustworthiness.
The research, published in PNAS Nexus, assessed perceptions using the Factors Assessing Science’s Self-Presentation (FASS) framework. This rubric evaluates credibility, prudence, bias, self-correction and perceived benefits. Compared to general science and climate science, AI fared poorly, particularly in the domain of prudence. Respondents expressed concern that AI researchers are insufficiently cautious, especially when it comes to preventing unintended consequences.
No improvement in AI perceptions over time
To determine whether public skepticism stemmed from unfamiliarity with the technology, the researchers examined changes in perception between 2024 and 2025. Despite AI becoming more visible in daily life during this period, attitudes remained largely unchanged. The persistence of negative views suggests that familiarity alone does not necessarily lead to greater public acceptance.
Climate science remains politically polarized; AI less so
Previous research has shown that political ideology strongly influences attitudes toward certain scientific fields. For example, climate science in the US has been shaped by partisan debates, and trust in medical science declined among Republicans after the COVID-19 pandemic. However, the study found that AI science is not currently subject to the same level of political division. Respondents across the political spectrum held relatively similar views about AI scientists, indicating that the field has yet to be heavily politicized.
Methodology and authorship
The findings are based on data from a national probability sample of US adults. The study was co-authored by Dror Walter, associate professor of digital communication at Georgia State University and APPC distinguished research fellow; Yotam Ophir, associate professor of communication at the University at Buffalo, State University of New York; Patrick E. Jamieson, director of APPC’s Annenberg Health and Risk Communication Institute; and Kathleen Hall Jamieson, director of the Annenberg Public Policy Center.
Reference: Walter D, Ophir Y, Jamieson PE, et al. Public perceptions of AI science and scientists relatively more negative but less politicized than general and climate science. PNAS Nexus. 2025. doi:10.1093/pnasnexus/pgaf163
This content includes text that has been generated with the assistance of AI. Technology Networks' AI policy can be found here.
