How Higher Ed Institutions Are Using Built-In Generative AI Tools

USF uses Microsoft 365 for email and intranet functions, and uses Microsoft Teams across both the academic and business spaces. Copilot is available as part of this technology package.

“One of the big things about having Copilot is that it’s available to every faculty and staff member and student at USF,” Henkel says. “We level the playing field immediately.”

In higher education, “people are going to use these things either way,” he says, and with Copilot built into existing Microsoft products, “they can use things that we know have data protection and privacy at the forefront.”

The USF technology leaders see a range of emerging use cases.

For educators, “it's hard to be relevant to our entire student body if I'm talking to 300 different students who all have different interests,” Henkel says. “AI now has the ability to create examples that are relevant to the individual but still grounded in the theory or context of the learning endeavor.”

DISCOVER: Microsoft’s CIO of Education addresses the impact of AI across campus.

On the business side, “it’s especially helpful for Microsoft Teams meetings — summarizing the meetings, getting tasks out of the meetings,” Fernandes says. “We’ve also noticed a lot of people using AI to accelerate their thought process, to start a business case or to begin summarizing documents that they’re preparing.”

Rice University Pilots Data Protection Project with Google Gemini

At Rice University, Vice President for IT and CIO Paul Padley calls the built-in tools a win-win. The university has campuswide Google and Microsoft contracts already in place.

“We get access to these tools without incurring extra costs,” he says. And because these companies are accustomed to working with the academic world, “they know about the Family Educational Rights and Privacy Act. They know about data protection in ways that a startup company, for example, might not.”

Rice already has a Gemini pilot running among a select group of students, with an eye toward “understanding what data is protected and how it’s being used,” Padley says. “We really have to be careful that things we’re using in an academic environment are FERPA-compliant. We’ve only recently gotten clarity on that for those tools in the past couple of months, and we’re still exploring what that means.”

Once the pilot is complete, Rice will scale up its efforts, according to Padley. This will require a series of steps before implementation.

“We would bring it to our AI advisory committee and give them another opportunity to chime in,” he says. “We may bring it to the faculty senate as well if it’s affecting pedagogy. We really need to go through those governance practices.”

READ MORE: Google Gemini has practical applications in higher education.

One early use case already has emerged. OpenStax, nonprofit initiative out of Rice, makes college textbooks available for free. A recent partnership between OpenStax and Google Gemini could potentially make them “even more easily available,” Padley says.

It may soon be possible, for example, for Gemini to leverage OpenStax content to inform faculty lectures, writing lessons that address student learning preferences.

“If I took the course content and the research on student learning and merged them with the AI capabilities, I might be able to produce content that would be better tuned to how the students are going to learn,” Padley says.

Built-In Generative AI Tools Offer Customizable Security Parameters

At Carnegie Mellon University, IT leaders like the built-in tools for their inherent data security.

With AI delivered by major vendors within existing products, the school can customize its protections.

“We can develop appropriate terms and conditions, from a contract standpoint, to protect our intellectual property and the private data and sensitive data that we’re the custodians of,” says Stan Waddell, vice president for IT and CIO. “That’s extremely important and helpful.”

When it comes to using those AI capabilities, “the No. 1 use case that I hear is simple productivity enhancement,” he says. “You need to write an email, and you use prompts to draft the email. Or you need to refine some writing or shift the voice of the writing, and you use prompts to do that versus going through numerous iterations on your own.”

In addition, generative AI is getting smarter, he says.

“The tools are getting better with financial analysis and math. Folks can do summaries of things that are more numerical,” he says. “If you have a fairly substantial spreadsheet or an article and you just want to cut to the chase, you can ask it to provide a prose summary, even if that spreadsheet or article has some financial or numerical information in it.”

With the emergence of built-in tools from Microsoft and Google — vendors already well-established in higher education — experts predict a rapid uptick in generative AI adoption.

“We’ll see impacts across institutions, from day-to-day operations, behind the scenes, all the way into teaching and learning applications,” says EDUCAUSE’s Robert. “Institutions are expecting to integrate generative AI more and more across every area of their organization.”