Universities are splitting into two camps: those rebuilding degree programs around AI and those adding AI tools on top of unchanged curricula.
Kogod, Arizona State, and Miami Dade show measurable gains from full redesign, including a 40% enrollment jump at Kogod.
Barnard, Michigan, and Columbia chose speed and flexibility, but governance questions are still catching up to deployment.
Universities around the world are grappling with the same fundamental question: What role should AI play in higher education? The answers are increasingly dividing institutions into two distinct approaches, and the gap between them may ultimately have a greater impact than any individual course or technology.
Some universities are redesigning entire curricula, retraining faculty, and integrating AI as a core component of their academic programs. Others are incorporating AI into existing structures through certificates, workshops, and digital tools while leaving their broader curricula largely unchanged. Neither approach is inherently superior; however, each offers a different educational experience, and the impact of these choices is becoming increasingly evident in student outcomes.
American University's Kogod School of Business offers the clearest case study. In early 2024, its leadership gave faculty six weeks to rethink what AI-literate business education should look like. What came out was not a bolt-on course but a full redesign. New AI Artisan classes cover foundational knowledge, while AI Sage courses embed hands-on AI work directly into finance, marketing, and accounting. Every incoming student now takes mandatory AI workshops.
The school also created an AI Instruction Faculty Fellow role to keep teaching current and to train instructors before students were expected to use the tools. Three years on, undergraduate enrollment has climbed 40%, applications are up 50%, and more than 90% of faculty now use AI in their teaching. Kogod rebuilt courses first, invested in faculty second, and taught AI's limitations before its strengths. The point was never just to teach students how to use AI. It was to teach them to question what it produces.
Arizona State University took a different route to a similar outcome. After partnering with OpenAI in 2024, ASU rolled out ChatGPT Enterprise campus-wide but let individual departments decide how to use it. The AI Innovation Challenge invited faculty and staff to pitch their own projects rather than follow a top-down plan.
Hundreds of proposals grew into more than 200 active initiatives spanning nearly every discipline, from research assistants to simulated patient conversations for health science students. By 2025, the partnership had expanded to ChatGPT Edu, backing more than 500 projects. ASU shows that large institutions can transform curricula without centralizing every decision, as long as shared tools and incentives exist.
Miami Dade College proves this kind of change doesn't require a big budget. Starting in 2021, the college trained hundreds of faculty, partnered with IBM on AI for All, and launched Florida's first associate's and bachelor's degrees in artificial intelligence. It later opened an AI center and is moving toward requiring every student to complete an applied AI course, with ethics built in from day one. Enrollment across its technology programs has hit record highs. The driver wasn't honesty but leadership commitment.
There is an interesting trend that can be observed in these institutions. Not all of the universities that move faster are wealthy, but all of the ones that move faster have governance structures that enable curriculum decisions to be made quickly.
Barnard College chose a lighter touch. Instead of redesigning majors, it focused on giving every student access to AI tools through a partnership with Google, which brought Gemini and NotebookLM to campus, paired with a four-level literacy framework. This model moves faster and disrupts less. However, it has also exposed a weakness. Faculty have raised concerns about privacy and oversight after the tools were already in students' hands, a reminder that governance built after deployment tends to lag behind the technology it's meant to manage.
The University of Michigan and Columbia University follow a similar pattern. Michigan offers short AI courses through its Centre for Academic Innovation. Columbia provides professional AI certificates alongside tools like CU-GPT. In both cases, AI sits beside the degree rather than inside it, treated the way universities once treated coding or data science: a useful add-on, not a requirement.
Also Read: Top AI Learning Programs Every Developer in India Should Consider in 2026
The overhaul of schools moved on to governance early. Kogod trained faculty before launching anything. Miami-Dade built ethics into required coursework from the outset. Barnard's governance questions, by contrast, surfaced only after the tools were already deployed.
Faculty adoption tells a similar story. Kogod tracks its numbers publicly and watched them rise from roughly half of the faculty to over 90% in a year. No comparable figures exist at the add-on schools, since participation there remains voluntary.
The risks differ too. Curriculum redesign takes more coordination but produces consistent outcomes across students. Add-on strategies are quicker to launch but risk leaving preparation uneven and governance fragmented.
The real divide in higher education isn't public versus private, or large versus small. It's between schools that treated AI as a reason to rebuild and schools that treated it as one more tool to fit into what already existed. Kogod, ASU, and Miami Dade can already point to enrollment gains and faculty adoption numbers.
Barnard, Michigan, and Columbia bet on speed and flexibility, and their long-term results are still unfolding. The universities that remain competitive may not be those that adopted AI first but those that built academic systems capable of adapting every time AI evolves.
Also Read: Top AI Courses in 2026 for Beginners Without a Tech Background
True learning is not the only thing happening in AI. A university that views technological change as a learning opportunity, not a software upgrade, will be able to produce students who are more likely to be able to meet the demands of a changing workforce with constant adaptability.
Top AI Engineering Colleges in India 2026 (Rankings & Courses)
Best Colleges Offering B.Tech in AI and Machine Learning in India
Rebuilding an AI curriculum integrates artificial intelligence across multiple disciplines, making AI literacy a core learning outcome. Adding AI courses introduces standalone electives, certificates, or workshops without fundamentally changing existing degree programs.
Institutions such as American University's Kogod School of Business, Arizona State University, and Miami Dade College have redesigned curricula to embed AI across academic programs, combining new courses, faculty training, and institution-wide AI initiatives.
Universities are redesigning curricula to prepare students for an AI-driven workforce. Employers increasingly expect graduates to understand AI tools, ethical considerations, and practical applications across business, healthcare, engineering, and other disciplines.
Standalone AI courses improve awareness and technical skills, but experts argue that integrating AI throughout the curriculum provides broader, discipline-specific experience that better reflects how AI is used in modern workplaces.
The case studies show that successful AI adoption requires institutional strategy, faculty development, ethical governance, and continuous curriculum updates. Universities that align AI with long-term academic goals are better positioned to prepare future-ready graduates.