Report Sets Out How Colleges Can Build Rigorous AI Programs Across Disciplines

Two models for combining AI with other fields

The report was developed through workshops that brought together 30 participants from 29 institutions. Its authors include Emmanuel S. Peters (AI4ALL), Douglas Fisher (Vanderbilt University), Leen-Kiat Soh (University of Nebraska-Lincoln) and Rebecca Allen (Mount St. Joseph University), along with Sri Yash Tadimalla and Noah Cowit of the Computing Research Association.

A starting point is a distinction between two models. An interdisciplinary AI program centers expertise in AI while incorporating knowledge, concepts and methodologies from another domain. An AI-augmented disciplinary program does the reverse: it centers expertise in a non-AI field while integrating AI literacy, tools and methodologies into that discipline. According to the report, the choice can shape everything from course requirements to the kinds of problems students learn to solve.

Whichever model is used, the report identifies common foundations. Institutions could develop a foundational technical AI course that also includes socio-technical content, plus an AI literacy course focused on responsible use. Programs should also ensure students engage with AI as a socio-technical system, covering privacy, bias, accountability, environmental impacts and other ethical, professional and societal questions. These topics could sit in a dedicated course or be spread across the curriculum.

Identity, capstones and curriculum design

The report also points to how students understand their place within their fields. Educators, it says, must help students consider not only how to use AI tools but also how AI changes the nature of disciplinary expertise, professional judgment and their sense of belonging. Soh, a professor at the University of Nebraska-Lincoln, said: "It's not just about what AI allows students to do, but how it changes the way they identify themselves as learners and future professionals."

Recommendations go beyond classroom courses. They include:

• interdisciplinary senior capstones involving real-world applications
• undergraduate research
• internships and co-ops
• an interdisciplinary AI minor open to both computing and non-computing students

On curriculum design, the report recommends looking for shared requirements that satisfy expectations in both AI and the partnering discipline, which would reduce redundancy and increase curricular capacity. Institutions could also consider bridging courses that introduce AI students to another discipline, or modular one-credit "AI labs" that give hands-on practice applying AI methods within other courses. Applied courses should draw on authentic problems, data and practices from the partnering discipline, and students should learn to communicate across disciplinary boundaries.

Structures that support faculty collaboration

The report states that courses alone will not create genuinely interdisciplinary programs. It calls on institutions to identify both administrative and academic champions for AI education, develop a shared understanding of AI literacy and fluency, and build cultures that support collaboration between AI-related and other academic programs. Suggested structures include co-teaching, cross-departmental appointments, seed grants, faculty incentives, revenue-sharing agreements and campus-wide communities of practice.

Roderick Lee of Pennsylvania State University said that interdisciplinary AI education "cannot rely on individual faculty members informally finding collaborators in other departments and taking on the work in addition to their existing responsibilities." Institutions, he said, must provide time, incentives, resources and administrative support.

Enrico Pontelli, Dean of the College of Arts and Sciences at New Mexico State University, said such education "should be a genuine partnership," with faculty from both areas having a meaningful role in shaping the curriculum.

What it means for academic leaders

The Computing Research Association describes the central message for academic leaders as follows: expanding AI education across disciplines is not simply a matter of adding AI content to existing courses. It requires decisions about curriculum, faculty collaboration, institutional structures, and what students should be able to do at the intersection of AI and another field.

The goal, according to the article, is to prepare students to understand AI, evaluate it, apply it responsibly and connect it with the knowledge and practices of their own disciplines. The report is one of seven NSF LEVEL UP AI workshop recommendation reports. NSF LEVEL UP AI is a collaborative effort involving ACM, CRA, IEEE-CS, AAAI and other partnering organizations, supported by the U.S. National Science Foundation under Grant No. CNS-2434416.