University of Michigan’s Information and Technology Services (U-M ITS) is now offering its suite of generative artificial intelligence (GenAI) tools to the campus community.

In a press release, U-M ITS said its generative AI tools emphasize the importance of equity, accessibility, and privacy. The school noted that all data shared with its AI services is private and will not be used to train AI models. U-M ITS also touted that its AI platform has been designed and tested for accessibility, noting that its GPT service works “seamlessly” with screen-readers.

“We passionately believe that everyone at U-M should have access to the most powerful technology tools available. At the core of our GenAI services is the commitment to provide tools and technologies that enhance, support and augment humanity,” said Ravi Pendse, U-M’s vice president for IT and chief information officer.

“Using these tools responsibly, I am confident that our community of scholars will make a positive difference in the world,” Pendse said. “I am so proud that we’re leading the development of these services because they will be a gamechanger for how colleges use GenAI going forward.”

In the press release, U-M said the ITS AI Services have several key differentiators from most other generative AI tools like ChatGPT, which charge monthly subscriptions for increased usage limits. The school noted that for most services, U-M will initially be providing no-cost levels of access with “generous” usage limits for students, faculty, and staff.

The school is initially launching three AI services that will be available to the university community across Ann Arbor, Flint, Dearborn and Michigan Medicine campuses.

The first new service – U-M GPT – lets users engage with popular GenAI models like ChatGPT and other large language models. U-M GPT will initially be provided at no cost to the U-M community.

The second service – U-M Maizey – will allow U-M users to access AI language models to query and question their own datasets. This tool can seamlessly connect to popular platforms like Google and Canvas, allowing it to extract valuable insights from user data.

The final of the initial AI offerings – U-M GPT Toolkit – will provide a platform to construct, train and host AI models securely and at scale. The school said the toolkit is optimized for advanced AI designers and offers a creative platform where users can tailor solutions to meet a wide variety of needs.

In terms of security, the AI platform has been approved for use with moderate sensitive data upon its release. Currently, using data that is above a moderate-sensitivity classification, including protected health information, is not allowed. However, U-M ITS said it is working to expand the data types that may be used within the GenAI tools in the coming months.

“Generative AI is here to stay,” said Robert Jones, ITS executive director of support services, who led the development team for the new AI platform. “U-M is taking this opportunity to lead the discussion and help show higher ed and the rest of our global community how you can provide fair, responsible and forward-thinking access to this kind of technology for an entire community.”

Read More About
Kate Polit
Kate Polit
Kate Polit is MeriTalk SLG's Assistant Copy & Production Editor, covering Cybersecurity, Education, Homeland Security, Veterans Affairs