There's an important article that's just come out (technically as an accepted pre-publication version) in the Harvard Data Science Review titled "Institutional Efforts to Help Academic Researchers Implement Generative AI in Research ". This uses the recent experience at the University of Michigan as a case study of institutional-level strategic planning to support the research enterprise in adopting generative AI as a new large-scale technology; there have been similar efforts at several other institutions, notably Harvard and the University of California, San Diego. At the same time, the authors suggest that we may be moving into an era where other large scale technologies will create similar opportunities across the research enterprise (efforts like the Cloud Labs at Carnegie Mellon University come to mind here). Very much worth reading. See

https://assets.pubpub.org/9evdl9pm/Liu%20&%20Jagadish%20(2024)_Just%20Accepted-71708981791796.pdf

For another view of the University of Michigan experience from the perspective of IT leadership, see

https://er.educause.edu/articles/2024/2/how-and-why-the-university-of-michigan-built-its-own-closed-generative-ai-tools

I would welcome links to detailed information about what other institutions are doing in the development of institutional resources. It seems likely that the recently announced NSF pilot National AI Research Resource effort will also open up some additional options in this area.

Echoing  this theme, I had an opportunity last November to a discussion in London with a number of major UK research universities on institutional strategies regarding AI. They are thinking about the issues here quite differently. Our colleagues at JISC, who hosted the meeting, very recently published this summary.

https://www.researchinformation.info/analysis-opinion/three-key-themes-artificial-intelligence

Clifford Lynch
Director, CNI