By: Ron Woody
Last October I attended the annual Educause Conference in San Antonio, Texas, with over 7,000 higher education IT professionals. As you can imagine with a conference that size, the number of breakout sessions offered during each block was overwhelming. As you might also imagine with AI in the news every day, it seemed that nearly half of the sessions offered included “GenAI” in their title.
In the handful of GenAI-related sessions I attended, it was quite clear that most campuses are wrestling with similar questions around this topic. What frameworks/policies/practices need to be in place to ensure the appropriate and effective use of GenAI on campus? How do we protect sensitive university data from being used by other organizations’ large language models (think Microsoft, Google, OpenAI, etc.)? How do we balance our faculty’s academic honesty concerns with the responsibility to teach our students to effectively use GenAI tools in their future lives and careers? These are all certainly among the questions and issues we’re discussing on our campus.
At EMU, we first focused on the question of protecting sensitive university data (including student and employee data). Key to this focus was providing a University-licensed alternative to individual free and paid accounts with tools like ChatGPT. While they are very powerful, individual accounts with tools like ChatGPT use any data provided to populate their proprietary large language models (LLMs), thus exposing the work products (and potentially protected data) of EMU users.
Last summer, IT made available to all EMU students and employees the Microsoft Copilot app. The Copilot chatbot app is included in the University’s Microsoft Campus Agreement licensing. The app can be accessed via the web (copilot.microsoft.com) or mobile apps available in the Apple App Store or Google Play Store. Once logged into the app with your EMU NetID credentials, Copilot provides similar functionality to tools like ChatGPT, but with the benefit of all data being securely retained in a private LLM not used outside of EMU accounts. Details about how to access and log into Copilot are available here.
A couple of important notes about the Copilot app:
Copilot chat sessions do not automatically save. Please review the article linked above for suggestions on how to retain copies of your prompts and Copilot responses
Our Microsoft licensing does not include the version of Copilot that can be added to Microsoft Office apps like Word and Excel
Understanding that the Copilot app doesn’t provide the full functionality of products like ChatGPT, such as the ability to create additional apps, faculty can request pre-approval to purchase licenses for other GenAI products using the Software & Cloud-based Service Purchase Pre-approval request form. These requests are reviewed for the academic or business use, to evaluate the need for a tool beyond Copilot and for the types of data which will be used.
Looking forward, IT is in the early stages of working with partners across campus to address the other questions mentioned above including how to protect sensitive data and developing frameworks/policies to ensure appropriate use of GenAI.
First, IT is planning updates to EMU’s Sensitive Data Guide. This guide allows students, faculty, and staff an “at-a-glance” review of the types of data that can be used and stored within various applications. Planned updates will address GenAI tools related to the various types of sensitive data commonly used at the university.
Second, IT is developing a policy related to the appropriate and effective use of GenAI tools. We have already engaged in early conversations with the IT Steering governance committee and have requested this as an agenda item for an upcoming Faculty Senate Technology Issues Committee meeting.
Finally, IT is managing small pilots for a few other GenAI tools for which EMU is licensed. These tools include:
Zoom’s AI Companion - This tool allows the automated creation of meeting transcripts and the ability to ask in-meeting questions like “catch me up” or “was my name mentioned” if you joined late, or to prompt for the creation of a to-do list of action items mentioned during the meeting. When added to a meeting, the tool sends a summary to the host once the meeting concludes.
Google’s Gemini app (formerly called Bard) - Similar to Microsoft’s Copilot App, Gemini provides both web-based and mobile app access for developing content based on prompts. Also similar, our Google license only provides access to the app and not to the add-on available for Google Docs.
Google’s NotebookLM - This tool allows people to gain insights from a variety of sources including uploaded files or documents (Google Docs, PDFs, txt files, and audio files), typed (or pasted) text, or links to web resources such as page URLs or Youtube videos.
Once you have added sources, you can then either provide a prompt question or use the built-in Studio tool to generate five types of automated responses: 1) a study guide, 2) a briefing doc, 3) an FAQ doc, 4) a timeline doc, or 5) a “Deep Dive Conversation”. The Deep Dive conversation tool produces an audio podcast with two GenAI-produced English-speaking hosts discussing the content you added. Some may call the resulting podcast “scary real.”
Much more information will be forthcoming throughout 2025 as policy development continues. Faculty who have questions or may be interested in participating in the Gemini or NotebookLM pilot groups should contact Ron Woody (rwoody@emich.edu) to discuss.
Ron Woody
Ron Woody serves as EMU's Chief Information Officer (CIO) overseeing the Division of Information Technology. He joined EMU IT in 2000, serving in various leadership positions and then assuming the CIO role in 2016. Ron is also an alum, graduating in 1986 with a B.S. in secondary education (Political Science/History).