The average undergraduate degree consists of 40 3-credit courses. Our general education program is 12 courses. At my university the current AI-enhanced majors require 10 AI-focused courses delivered by an external department. At that scale, this is not integration; it’s curricular restructuring.

Every AI-enhanced major gets the same AI curriculum, regardless of discipline. And at ten total courses, It’s more like a second general education curriculum, but one that is AI-focused. This strikes me as overkill for most students and programs.

Perhaps more concerning is how such a program would shape the discipline and department being “enhanced.” Inevitably the gravitational center of that department would be shifted by the terms of another. In my view, a defensible curricular model for AI enhancement would be disciplinary-specific and three courses rather than ten. This curriculum would include:

1. Explore
How is artificial intelligence addressed, studied, and deployed within our discipline? What research methods intersect with it? Where does it already operate in professional practice?

2. Critique
What disciplinary concerns do we have? What broader social, political, environmental, aesthetic, and epistemological critiques apply? How do existing theoretical traditions in the field illuminate AI?

3. Practice
Given both the possibilities and the concerns, how might students engage AI in discipline-specific ways? What does responsible, situated experimentation look like?

This model is modest, coherent, and proportionate to the structure of a forty-course degree. It embeds AI inside disciplinary logics rather than treating it as an imported technical specialty.

The prevailing rhetoric surrounding AI in higher education insists that it is unprecedented. It is not.

In Media Study, for example, we have been teaching the study of software, infrastructure, algorithms, automation, and computation for over fifteen years. Fields such as media archaeology, software studies, critical code studies, science and technology studies, platform studies, and infrastructure studies have long examined the material and cultural dimensions of computational systems. These approaches sit alongside Marxist, feminist, decolonial, gender studies, and environmental critiques of technological systems, among others. Our students are also encountering AI in their production curriculum: Adobe and other software, as well as in the operation of cameras and related hardware. They encounter the issue in programming (social media, games, etc.) and film editing courses.

Media Study is not alone in its situation. All of the humanities have been exposed to computation for more than 50 years. AI isn’t new in this sense. For example, the humanities have digital humanities, textual encoding, history of technology, cybernetics, and algorithmic culture. What is new is the speed of institutional uptake and the concentration of capital behind “AI” as a technical investment and a branding opportunity.

Developing and delivering a three-course sequence would be workable for most departments. The more challenging part may be developing a department culture that welcomed such work as it overflowed into the major, as it should if our students are using it. As hard as that transition might be, it would certainly be better than having students take 10 classes in some other department and have that somehow become part of your major.

Universities are under pressure to brand themselves as AI leaders. But branding is not curriculum. Liberal arts disciplines should not allow institutional urgency to overwrite their intellectual commitments and history. AI integration should be disciplined, proportional, and grounded in the epistemologies of disciplines—not imported wholesale under the banner of innovation.