The Pulse of Higher Ed

When we reflected on the first three weeks of the UPCEA AI Study Groups in early June, one theme had already become clear: AI implementation is not simply a technology challenge. It is a leadership challenge. 

By the end of the six-week pilot, that theme had deepened. Across two cohorts of UPCEA members, participants moved from broad discussion about AI experimentation, governance, and institutional readiness, into more focused conversations about use cases, roadmaps, change management, student support, faculty development, and the practical work required to move AI from possibility to responsible implementation. 

The Study Groups were designed to help members build community and engagement around AI at their institutions. They also served as a listening and learning space for UPCEA as members continue to navigate fast-moving questions about AI strategy, policy, operations, teaching and learning, and workforce readiness. 

The final weeks reinforced a core insight: institutions do not need to have every answer before they begin. But they do need intentional structures, realistic expectations, and communities of practice that help them learn as they go. 

AI Implementation Is Moving from Curiosity to Infrastructure 

The early weeks of the Study Groups focused heavily on where institutions are in their AI journeys. Some participants described early-stage experimentation, while others shared more mature approaches involving AI councils, task forces, enterprise tools, faculty development programs, AI literacy initiatives, and operational pilots. 

By the final weeks, the conversation had shifted from “What are institutions doing?” to “How do we make this work responsibly and sustainably?” 

Participants discussed AI as part of institutional infrastructure, not just as a set of tools. Successful implementation requires attention to data systems, privacy, procurement, cybersecurity, workflow mapping, staff capacity, faculty engagement, student communication, and governance. Several participants noted that the most difficult barriers are not always technical. They are often cultural, organizational, and procedural. 

This matters especially for online and professional continuing education units, which often operate at the intersection of academic innovation, adult learner support, workforce alignment, enrollment strategy, and operational agility. For these units, AI is not an abstract future issue. It is already shaping advising, marketing, course design, student communication, curriculum planning, employer engagement, and administrative workflows. 

Practical Use Cases Helped Move the Conversation Forward 

The second half of the Study Groups featured practical institutional examples that helped participants think more concretely about implementation. 

One student support pilot demonstrated how AI can help staff consolidate data from multiple systems, identify students who may need timely outreach, and personalize communications at scale. The pilot significantly increased proactive outreach while reducing the time needed to prepare individual student communications. Just as importantly, the example surfaced the implementation realities behind the results: data integration, workflow mapping, system testing, staff trust, transparency, vendor partnership, and return-on-investment conversations with institutional leadership. 

Another institutional example focused on faculty development and teaching and learning. Participants explored how AI can be introduced through workshops, faculty cohorts, mini-grants, instructional design support, AI literacy modules, and learning management system tools. The discussion also addressed persistent concerns about academic integrity, privacy, regular and substantive interaction, student comfort levels, and the importance of ensuring that AI amplifies rather than replaces human teaching and support. 

Together, these examples helped participants see that promising AI work does not begin with sweeping transformation. It often begins with a specific problem, a defined audience, a manageable pilot, and enough structure to learn from the effort. 

“Crawl, Walk, Run” Remains a Useful Discipline 

One of the most resonant frameworks across the Study Groups was the “crawl, walk, run” approach. Participants repeatedly returned to the importance of starting with a contained use case, building trust, demonstrating value, and scaling only after the institution understands what is working and what is not. 

The distinction between what is possible, plausible, and probable also helped participants frame conversations for different audiences. Senior leaders may need to understand the broad possibilities of AI, but end users often need clarity about what is realistic, safe, and immediately useful. A project that sounds simple in theory may require significant process documentation, data review, testing, training, and communication before it can be implemented responsibly. 

This was especially evident in conversations about AI agents. Participants were interested in agents for advising, financial aid communication, reporting, student outreach, and internal process support. But the Study Groups also highlighted important questions: Who is accountable for an agent’s output? How will students know when AI is being used? What policies govern the use of real images, voices, or institutional branding? What data can be used safely? What human oversight is required? 

The lesson was not to avoid these tools. It was to implement them with discipline. 

Governance Needs to Enable, Not Paralyze 

Governance was a recurring topic throughout all six weeks. Participants described a range of institutional approaches, including AI committees, executive councils, task forces, working groups, acceptable use policies, data governance structures, procurement reviews, and faculty-led guidance. 

Yet many institutions are still operating in a decentralized environment. Some participants noted unclear directives, uneven communication, leadership transitions, cybersecurity concerns, and inconsistent faculty messaging to students. Others described progress through cross-functional groups, grassroots communities of practice, and unit-level leadership even when institution-wide policy was still emerging. 

A key takeaway is that governance should not become a reason for inaction. Institutions need guardrails, but they also need ways to learn. The most promising approaches appear to combine broad principles with practical pathways for pilots, review, training, and iteration. 

Participants also emphasized that AI strategy should not be disconnected from digital strategy. Institutions need to understand how AI fits into broader technology, data, learner success, workforce, and academic innovation goals. In the strongest examples, AI was not treated as a side project. It was connected to mission, strategy, and institutional capacity. 

AI Literacy Is Becoming a Shared Responsibility 

The Study Groups made clear that AI literacy is no longer relevant only to technologists, faculty innovators, or instructional designers. It is becoming a shared responsibility across the institution. 

Participants identified many roles affected by AI, including faculty, students, instructional designers, advisors, IT staff, administrative assistants, marketing and enrollment professionals, financial aid staff, curriculum leaders, and senior administrators. Each group needs different kinds of support. 

For some, AI literacy begins with basic understanding: what AI can and cannot do, how to prompt effectively, how to recognize bias or inaccuracy, and how to use approved tools safely. For others, the need is moving toward AI fluency: understanding how to move across tools, apply AI in context, evaluate outputs, design workflows, and use AI as a “power tool” while maintaining human judgment. 

This distinction between literacy and fluency became particularly important in conversations about workforce preparation and professional education. Students need guidance not only about academic integrity, but also about how AI is changing their fields, their future work, and the expectations of employers. Institutions need to help learners use AI ethically, critically, and effectively. 

The Human Side Is Still the Hardest Part 

Across both cohorts, the most consistent implementation challenge was not tool selection. It was people. 

Participants discussed faculty fatigue, student anxiety, staff concerns about job displacement, uneven access to tools, lack of time, inconsistent policies, and the challenge of building buy-in across departments. They also shared the importance of framing AI as a way to enhance human work rather than erase it. 

This framing matters. In the most compelling examples, AI was used to help humans do more meaningful work: advisors could spend less time compiling information and more time supporting students; faculty could experiment with new forms of engagement; staff could reduce repetitive tasks and focus on judgment, strategy, and relationship-building. 

But that outcome is not automatic. It requires intentional communication, training, transparency, and trust-building. Institutions must acknowledge fear and uncertainty while also creating space for experimentation and learning. 

Peer Learning Was the Greatest Value 

The participant feedback reinforced the value of community. In the post-pilot survey, 80% of respondents rated the AI Study Group as valuable or very valuable. When asked what was most valuable, participants repeatedly pointed to learning from peers, hearing how other institutions are approaching AI, seeing examples of use cases, and understanding that others are wrestling with similar questions. 

Respondents also reported a meaningful increase in confidence. Before the Study Group, fewer than half of respondents rated their confidence as somewhat high or very high in feeling equipped to advance AI work or build a plan. After the Study Group, 90% rated their confidence as somewhat high or very high. 

The survey also showed that participants left with forward momentum. Forty-five percent reported having a specific AI workflow or use case to implement, while 55% said they had a direction and next steps that needed further refinement. In other words, every respondent left with either a concrete implementation idea or a clearer path forward. 

The UPCEA AI Hub as a Continuing Resource 

The AI Study Groups are part of UPCEA’s broader effort to support responsible, practical AI adoption in online and professional continuing education. The UPCEA AI Hub serves as a central resource for members seeking guidance, tools, research, use cases, and peer learning related to AI implementation. 

The next release of the AI Hub later this month will add more resources and tools to support member institutions. The release will include new and reorganized content aligned around foundational elements, applied areas, and transformation, as well as institutional case studies that highlight practical examples of AI implementation. Members will also see additional resources connected to AI readiness, maturity, governance, use cases, and operational planning. 

The Study Groups helped surface what members need most: examples they can learn from, frameworks they can adapt, tools they can use with colleagues, and opportunities to connect with peers who are asking similar questions. 

Final Takeaways 

After six weeks, several final insights stand out: 

• AI implementation must be mission-aligned. The goal is not to adopt AI for its own sake, but to improve learning, support students, strengthen operations, and prepare institutions and learners for a changing world. 

• Institutions should start with real problems. The strongest use cases begin with a clear challenge, defined users, and a manageable scope. 

• Governance should support responsible action. Institutions need policies and guardrails, but they also need room to pilot, evaluate, and iterate. 

• AI literacy must be broad and role specific. Faculty, staff, students, and leaders all need support, but not the same support. 

• Human trust is essential. AI adoption depends on transparency, communication, training, and a clear message that AI should enhance human work and judgment. 

• Peer learning accelerates progress. Members benefit from hearing what others are trying, what barriers they face, and what lessons they have learned. 

The Study Groups confirmed that UPCEA members are not waiting passively for AI to reshape higher education. They are asking hard questions, testing new approaches, building frameworks, and looking for ways to lead responsibly. 

The work ahead will require continued experimentation, shared learning, and practical tools. Through the AI Hub, CORe, future Study Groups, and member-contributed case studies, UPCEA will continue helping institutions move from AI exploration to action—grounded in the needs of adult learners, professional education, workforce alignment, and the future of online and professional continuing education. 

Thank you to the guest speakers, Eric McGee, Strategic Consultant for Membership Growth and AI Engagement with UPCEA, Jessica Salata, Director, Student Success for Arizona Online, Caleb Simmons, Vice Provost for Arizona Online, Heather Breittholz, Lead Instructional Developer at UConn eCampus, and Desmond McCaffrey, Interim Director of the Center for Excellence in Teaching and Learning (CETL) at the University of Connecticut for sharing their insights and expertise. And thank you to all of the participants for their active engagement in the study group throughout the six weeks. 

The AI Study Groups were facilitated by Vickie Cook, Senior Fellow and Strategic Advisor for UPCEA, with support from Stacy Chiaramonte, Senior Vice President Operations and Strategy for Research and Consulting, Mel Edwards, Director of Membership and Business Development, and Kathleen Ives, Chief Business Development Officer and Senior Vice President of Member Engagement.  

 

Content for this resource was refined with the assistance of AI. All text has been thoroughly reviewed, edited, and approved by UPCEA staff with subject matter expertise. References and links have been verified for accuracy and reliability.