As classrooms continue to integrate artificial intelligence tools across campus, Penn professors spoke to The Daily Pennsylvanian about how they are adopting the technology without compromising the learning process.
Penn first announced its university-wide AI policy in November 2023, which outlined formal guidelines for the use of generative AI and machine learning tools. While the policy provides a general framework for the Penn community, faculty across the University have taken varying approaches to using AI in learning.
“Penn embraces innovations such as generative artificial intelligence (“AI”) models in teaching, learning, research, and the effective management of Penn resources,” Penn’s guidelines state. The policy encouraged “transparency” and “accountability” and required students to follow standards set by courses when employing AI in the classroom.
In fall 2024, Penn launched the Penn AI Council – as part of the University’s 2023 “In Principle and Practice” framework – to provide “strategic input, coordinate interdisciplinary research, and catalyze new initiatives for Penn’s growing leadership in AI.”
“The AI Council was established to develop a strategy to coordinate AI activities within the university,” Professor René Vidal of the Perelman School of Medicine, who is co-chair of the council, told DP.
“We align schools, faculty, donors and alumni on a small set of talking points that can guide the University’s investments in terms of faculty we recruit or areas we support to obtain research grants,” he added.
Vidal clarified that the board “does not impose rules on anyone” or make specific recommendations to faculty and students.
“We really want Penn students to be better prepared for the ‘AI world’ and we want Penn to be the leader in training our students and faculty,” said natural sciences professor Bhuvnesh Jain – also co-chair of the AI Council.
“There are ways that AI can accelerate learning, and there are things about AI that you need to learn to do it well on your own,” he continued, saying instructors should teach “more innovative ways” of approaching AI.
Last semester, Jain taught a half-credit pilot course titled “Introduction to AI: Concepts, Applications and Impact. »
“I want students to really understand how big language models work and how they can use AI for their own benefit,” Jain said.
The course brought together 10 guest speakers — including venture capitalists, media experts and Penn faculty from multiple academic disciplines — to “expose students to Penn’s full range of expertise and a cross-section of professions” related to AI.
Other Penn faculty members shared how they are integrating AI into their classrooms.
“I used my favorite AI… to help me write a linear algebra textbook,” Associate Dean for Undergraduate Studies in the School of Engineering and Applied Sciences and professor of mathematics Robert Ghrist wrote to the DP. “This was a collaboration between me and the AI…it took me approximately 120 hours of work and over 55 days to produce.”
Ghrist also creates an AI “virtual course assistant” for each course he teaches. According to Ghrist, these virtual assistants incorporate “the curriculum, text, learning objectives, grading, etc., as well as instructions on how to help students study the material.”
When a student accesses one of Ghrist’s course assistants, they receive four prompts: “help”, “introduction”, “think”, and “leave”.
When the “introduction” prompt is clicked, the algorithm asks the student about their major and area of study, as well as what they would like to learn or review for the Ghrist course.
“[I] I spend a lot of effort thinking about how to use AI responsibly and creatively to help students take charge of their learning,” Ghrist wrote.
Other Penn professors have taken similar approaches to using AI. In Penn’s Japanese language program, for example, instructors use AI to enrich instruction and aid language retention.
“As a language educator, I do not view AI negatively,” Penn professor Megumu Tamura wrote in a statement to DP. “I find it very useful for thinking, generating ideas, generating visuals, and helping students locate appropriate reading materials more effectively.”
Although Tamura expresses optimism about AI, she adds that her view is not entirely uncritical.
“My concern is not whether AI works, but what kind of learning it produces,” Tamura wrote. “AI can be a powerful tool, but only if we design its use in a way that preserves the cognitive processes essential for true language growth. »
Tamura — alongside Japanese professors Ryo Nakayama, Saki Hirozane and Nana Kolb — presented a workshop at the Penn Language Center’s 2025 Language Educators Symposium titled “Full Transparency: Connecting Students, Instructors and AI.” The conference concentrate on ways to integrate AI technologies into language teaching.
“We integrated personalized GPTs into our course through an ‘integrated approach’ combining course-specific GPT-based tasks, task-based learning activities, and gamification elements,” Kolb wrote to the DP.
Kolb explained that Japanese language classes use AI in three ways: for interactive speaking and listening practice, interview tasks, and practice typing and character recognition.
“We believe AI complements human teaching by providing scalable practice opportunities and personalized feedback,” he added.
Kolb explained his belief that AI should complement traditional classroom learning, writing: “Successful AI integration requires intentional design that empowers rather than replaces student action. »
“It’s not difficult for a professor or even the entire computer science department to grab hold of AI and try to do great things, but it’s very difficult for a university to achieve this,” Jain concluded. “By starting by working as a team across all schools, I think we are in a pretty unique position to do that. »
Senior reporter Jack Guerin contributed reporting.
Journalist José Carlos Serrano contributes to data and business reporting and can be reached at [email protected]. At Penn, he studied English and political science.
