Amid the rise of artificial intelligence in classrooms at the University of Michigan, the University released the final draft of its AI Guiding Principles — a document containing six principles providing guidance on the use of AI in academic situations — Oct. 21.
The six principles consist of championing responsible, human centric AI; advancing the institutional mission; supporting critical thinking development while building AI literacy; leading with purpose and engaging in continuous improvement; centering ethics, integrity, and transparency; and ensuring privacy of data and of the University community.
The AI in Education Working Group initially investigated the use of AI at the University and shared their findings April 30 to shape how AI is utilized within the institution. The first draft of the University’s principles was published Sept. 3, and students and faculty members were invited to provide feedback.
In an interview with The Michigan Daily, Mika LaVaque-Manty, professor of political science and chair of the working group, said the group recommended against a singular policy, arguing it would not take into account the various needs of departments across campus.
“We had heard from students, faculty and staff that they felt lost in the sea of AI on what the University’s position was,” LaVaque-Manty said. “We very strongly felt after our study — after we studied this issue — that a university-wide policy would be a very bad idea, partly because it would be just too general and not specific and vague.”
In an interview with The Daily, Business graduate student Rounak Banik, co-president and founder of the AI and Emerging Tech Club, said the club worked with the Ross School of Business to develop its feedback on the AI guidelines. Banik said he was pleased to see a lot of the principles the group promoted in the final copy of the guiding principles.
“A lot of the things that we spoke about, I’m very happy that … it has figured in the guiding AI principles that was published by the University — mainly to do with having a human-centric approach,” Banik said.
The document is not a policy — instead, it is considered a “guiding foundation.” LaVaque-Manty said faculty and instructors still have ultimate say over the conduct of AI in their classroom.
“We recommend it — and I believe that the University also agrees — that ultimately, each faculty member should have a say on the implementation of AI in their classroom,” LaVaque-Manty said. “At the same time … we think that the best level for decision-making would be something like an academic department or any unit that controls curriculum.”
In an interview with The Daily, Business sophomore Julia Lotz, president of the AI Business Group, said she believes it is vital for faculty members to become just as literate in AI as the students while the new technology advances.
“It’s just as important — arguably more important — for the teachers and the faculty to understand what’s going on with AI,” Lotz said. “It’s super, super encouraging and amazing to see the (instructors) that really embrace AI. I, in the past year, have been forced to use it in my assignments, which in high school I would have thought never would happen in college.”
Banik said he believes the University is equipped to lead AI advancements, although he thinks the institution should prioritize developing AI from a human-centered perspective — the first principle released in the University’s guide.
“It is important that the conversation be centered around humans, and the conversation be centered around empowerment,” Banik said. “I feel like the University of Michigan is in a very good place to actually take the mantle over there to champion human-centric AI that preserves our critical thinking, that preserves our creativity, that preserves our humanity.”
The University of Michigan was the first university to provide a custom suite of generative AI tools, including U-M GPT, U-M Maizy and U-M GPT ToolKit. LaVaque-Manty said this provides an opportunity for students to explore the world of AI.
“I think college is about exploration — I would encourage students to try any of the tools that they are interested in, whether they like AI, whether they’re critical or whether they’re enthusiastic,” LaVaque-Manty said.
Lotz said she hopes the University will continually update its principles for AI as the technology advances.
“I would say my one hope is that they’re consistently updated,” Lotz said. “A year from now — not even a year from now, with the rate we’re at, with everything going on, it could be next week — something comes out, a new model comes out that has a whole other realm of implications that you then need to take into review.”
Daily News Contributor Shaena Foley can be reached out to at snfoley@umich.edu.
