Over the past six months, through the Carnegie AI Collaborative and our work with presidents and leadership teams across higher education, we have had a front-row seat to how institutions are navigating AI.
At the start, many of our conversations with presidents began with a question: “What should we be doing with AI?” Today, we think the better question is: “What kind of institution do we need to become to take advantage of it?”
After six months, the biggest lesson is that successful AI transformation isn’t primarily about moving faster—it is about building the institutional capacity to move deliberately at scale.
The presidents who seem best positioned aren’t necessarily the ones with the most pilots; they’re the ones connecting leadership + people + data + governance + institutional priorities + execution. Moving from isolated AI pilots to institution-wide adoption requires strength across all six of these areas.
AI may be the catalyst, but the work is fundamentally institutional, and fundamentally human. That is ultimately what AI readiness requires. Here are six lessons we are carrying forward from those conversations.
1. AI transformation is a human challenge before it is a technology challenge
The biggest barriers to AI adoption are rarely about access to the technology. They are about fear, uncertainty, trust, confidence, and culture. Faculty and staff are asking what AI means for their work, their expertise, their students, and the institution. Students are asking how to navigate differing faculty policies, while trying to develop literacy for career readiness and gaining core critical thinking skills.
The institutions making the most progress are investing as deliberately in engagement, training, communication, and change management as they are in technology.
The implication here is that you can buy AI tools, but you can’t buy adoption.
2. Your AI strategy will only be as strong as your data governance
AI has exposed weaknesses that institutions have tolerated for years: fragmented systems, inconsistent definitions, inaccessible data, unclear ownership, and uneven governance.
Before institutions can unlock AI at scale, they need to address three fundamentals: clean data, connected data, and governed data.
There’s also a growing frustration with the idea of handing institutional data to vendors only to have insights derived from that data effectively sold back to the institution.
Data ownership, portability, privacy, and vendor governance are becoming strategic, not just technical, questions. Strong AI governance starts with clarity around how data is managed, who makes decisions, and how institutions establish accountability as adoption expands.
3. Everyone thinks they are behind. Almost no one is as far ahead as everyone else assumes.
One of the most consistent things we hear from presidents is some version of: “Are we behind?”
The reality is that institutions vary enormously in experimentation, but very few have moved from experimentation to a coherent, institution-wide strategy. That should create some urgency—but also reassurance. There is still an opportunity to be deliberate rather than reactive.
The challenge now is turning what institutions learn from individual pilots into the leadership alignment, governance, infrastructure, and execution required for institution-wide adoption. For leaders trying to understand what that shift requires, our earlier article on AI readiness in higher education leadership looks more closely at the leadership, alignment, and organizational foundations needed to move forward.
4. The best AI strategies don’t start with AI. They start with institutional strategy.
The question shouldn’t be, “Where can we use AI?” It should be, “What are the most important problems our institution needs to solve—and where can AI materially accelerate our progress?”
Enrollment. Retention. Student experience. Research productivity. Administrative efficiency. Academic innovation.
AI can help advance each of these priorities. Its value becomes much clearer when it is connected to the institution’s strategy rather than treated as a parallel collection of pilots searching for a purpose.
An effective AI strategy should therefore begin with institutional priorities, not with a list of potential AI use cases.
5. Leadership and ownership matter, but there is no single right organizational model
Nearly every institution is wrestling with the same question: Who owns AI? Is it the CIO? The provost? The president? A new chief AI officer? A cross-functional committee?
Our experience suggests the title matters less than establishing clear executive sponsorship, accountability, decision rights, and cross-functional leadership. Leadership alignment also creates the foundation for setting priorities, allocating resources, and determining which opportunities move forward.
AI touches too many parts of an institution to live entirely within one office. At the same time, distributed ownership without clear leadership can quickly become fragmented experimentation.
Successful transformation requires broad participation and clear accountability. Whether AI leadership is centralized, distributed, or shared across a cross-functional group, those fundamentals matter more than the organizational chart itself.
6. Start with the boring stuff
Some of the most transformative opportunities are also the least flashy.
Before jumping immediately to “AI in the classroom,” institutions can create enormous value by tackling the foundational and operational work: data infrastructure, governance, policies, administrative workflows, knowledge management, repetitive processes, and staff productivity.
Doing the “boring” work creates tangible wins, builds institutional confidence, develops AI fluency, and establishes the infrastructure needed for more ambitious applications later.
A More Intentional Path Forward
Taken together, these lessons point to a broader conclusion: AI readiness depends on the institutional capacity to make thoughtful decisions, align around priorities, and execute with purpose. It means building capacity across leadership, people, data, governance, strategy, and execution so institutions can move from experimentation to adoption at scale.
Students have to remain central to that work.
AI will change how institutions operate, how faculty teach, how staff work, and how students learn and prepare for their careers. Progress should be measured by whether AI helps institutions strengthen the student experience, expand opportunity, improve outcomes, and preserve the distinctly human value of higher education.
The institutions best positioned for this moment are building the ability to learn from experimentation and adapt as the technology continues to evolve. The lasting advantage will come from the institutional capacity built around AI, not from any single tool or use case.
For institutions ready to move from experimentation to a more intentional approach, Carnegie can help.
Our AI Readiness Audit assesses current capabilities across leadership, people, data, governance, strategy, and execution, helping institutions identify priorities and build a clearer path forward.
Learn more about the AI Readiness Audit.
What AI Readiness Looks Like in Practice
What is AI readiness in higher education?
AI readiness is an institution’s ability to adopt and use AI in a purposeful, responsible, and sustainable way. It includes more than technology. Leadership alignment, workforce confidence, data quality, governance, institutional priorities, and execution all shape whether AI can create meaningful value. Together, these capabilities determine whether an institution can move beyond individual AI experiments and adopt AI effectively at scale.
How can colleges and universities assess their AI readiness?
Institutions can begin by looking across several areas: leadership and ownership, faculty and staff readiness, data infrastructure, governance and policy, strategic alignment, and the ability to move from experimentation to execution. The goal is to identify where strong foundations already exist and where gaps could limit progress.
Who should own AI strategy at a college or university?
There is no single organizational model that works for every institution. Effective approaches typically include clear executive sponsorship, defined decision rights, cross-functional participation, and clear accountability. Because AI affects academic, administrative, technology, and student-facing work, coordination across the institution is essential. The specific governance model matters less than ensuring that ownership, authority, and accountability are clear.
Why is data governance important for AI?
AI depends on reliable, accessible, and well-governed data. Fragmented systems, inconsistent definitions, unclear ownership, and weak privacy or vendor policies can limit the value institutions gain from AI. Strong data governance creates a more dependable foundation for responsible adoption at scale.
Where should higher education institutions start with AI?
Start with institutional priorities. Rather than asking where AI can be used, identify the problems the institution most needs to solve and determine where AI can meaningfully accelerate progress. Early opportunities may include administrative workflows, knowledge management, staff productivity, data infrastructure, and other foundational work.
How can institutions keep students at the center of AI transformation?
Institutions can evaluate AI decisions based on their impact on the student experience, access, learning, career readiness, and outcomes. As AI changes how institutions operate and how students learn, leaders have an opportunity to ensure the technology supports both institutional effectiveness and the distinctly human value of higher education.
What is an AI Readiness Audit?
An AI Readiness Audit gives institutional leaders a clearer picture of their current capabilities, gaps, and priorities for AI adoption. Carnegie’s approach examines areas including leadership, people, data, governance, strategy, and execution to help institutions identify practical next steps and move forward with greater clarity.
