Funding influences how scientists use AI models in research
Open Science, Closed Models: How Funding Shapes AI in Science
Digital Libraries
Summary
This study looks at how the source of money for scientific research affects the use of large AI models. It finds that public funding is linked to more open and shared AI tools, while private funding tends to support the use or enhancement of closed, proprietary AI models. Collaboration with industry and access to cloud computing credits often push research toward closed AI systems. The study also points out that researchers in less wealthy countries are mostly left out of these AI collaborations. Overall, changes in AI funding may shift scientific work towards more closed and proprietary AI infrastructure.
What this means in practice
- •For science funders: Adjust funding strategies to balance open and closed AI model use based on how funding source influences AI engagement in scientific research.
- •For cloud service providers: Design cloud credit programs that consider their influence on AI model openness in funded scientific projects to better guide research direction.
Authors
Ana Trišović, Janakan Sivaloganathan
Abstract
How funding shapes AI engagement in science is poorly understood despite its structural importance. We analyze 104,226 scientific papers (2018-2025), linking funding acknowledgments to how each paper engages with foundation models: whether it extends a model (fine-tunes or builds on it), uses one without modification, or references models only peripherally. Three findings emerge. First, funding source is associated with the character of engagement: public funding is associated with higher open-weight model engage- ment; private-only funding selectively enables extension with no robust openness association; mixed funding shows nominally the highest open-weight engagement rates in four of the seven largest disciplines. Second, papers acknowledging industry cloud credits are less likely to use open-weight models, consistent with credit programs steering research toward closed models. Third, industry collaboration carries an independent extension premium consistent with internal corporate resources flowing through coauthorship channels invisible to acknowledgment-based measures. This nexus concentrates in Computer Science and Global North collaborations; Global South research is largely excluded. As public funding contracts and industry compute provision expands, scientific AI work shifts toward closed proprietary infrastructure.