The U-M Generative Artificial Intelligence Advisory (GAIA) Committee report was created by faculty, staff, and students and "lays a proposed foundation for how U-M might live and work with GenAI." The report is not comprehensive but provides the university's approach to GenAI needs.
University of Michigan offers generative AI platforms to all active U-M faculty, staff, and students on the Ann Arbor, Flint, and Dearborn campuses - and to Michigan Medicine. U-M has licensed access to both externally created and hosted tools and those developed and hosted internally. Internal AI tools for the U-M community include examples such as U-M GPT and U-M Maizey (which we primarily focus on here) as well as many others. There are new tools still in development, and the tools already listed continue to be refined. You will want to check official U-M sources for the newest updates.
U-M's internal AI services benefit our community in many ways. Some of the most important reasons for investing in these have been to offer local supports and controls in areas such as equity, privacy, security, sustainability, and accessibility. U-M does not use uploaded content or queries as training data, and does not share content or queries for human-review.
Privacy of content is especially important in the medical environment, both for HIPAA as well as research data and original publications. For those concerned about personal privacy, pay special attention to the "How We Use Personal Information" section of the privacy notice linked above. Note that the statement "Provide persistence of user prompt history across sessions with the ITS AI services" is under user control, and that the user has the choice to delete their session history at the end of each session.