Field diffusion · General science · Social science
Study finds collaboration networks shape AI adoption in science
A peer-reviewed OpenAlex analysis identifies social, institutional, and technical correlates of AI adoption across scientific research.
Summary
Bianchini, Müller, and Pelletier analyze OpenAlex publication data to identify correlates of AI adoption and reuse in scientific research. They find that adoption is strongly associated with AI-rich collaboration networks and early-career collaborators; after large language models diffuse, computing access and institutional advantages appear less decisive, although social ties remain important.
AI role
AI is measured as a research method whose adoption and reuse are compared across scientists, collaboration networks, institutions, and fields.
Narrative role
This adds field-diffusion evidence about the conditions under which AI becomes embedded in research practice, showing that access to capable tools alone does not determine uptake and that collaboration networks may shape where acceleration reaches first.
Caveat
The observational bibliometric design identifies correlates of recorded AI use rather than causal effects of AI adoption, and publication-based measures can miss unreported tool use.