Higher education institutions are moving fast on AI, but “fast” looks different depending on where you stand. A ten-campus system has different constraints than a single liberal arts college. A research university with more than 150 faculty actively exploring AI has different needs than one that is still defining its approach.
At this year’s EDUCAUSE conference, Dr. Cristi Ford, Chief Learning Officer at D2L, will moderate a conversation with Dr. Sherri Braxton Castanzo (Bowdoin College), Dr. Gloria Niles (University of Hawaiʻi System) and Dr. Melissa Vito (University of Texas at San Antonio). Together, they’ll explore how institutional mission, culture, governance structures and community needs shape AI strategy in ways that no universal framework can.
Ahead of that discussion, we asked each panelist a simple question: what does your mission require that a generic AI policy wouldn’t capture?
Governance Through Dialogue
At Bowdoin College, a small liberal arts institution, AI governance has taken shape less as a policy document and more as an ongoing dialogue.
“We’ve been able to approach AI governance as a highly engaged, thoughtful community conversation rather than primarily as a compliance exercise,” says Dr. Sherri Braxton Castanzo, Deputy CIO for Digital Innovation at Bowdoin. “Our scale allows us to bring faculty, students and staff into the same conversations and adjust quickly as needs emerge. Rather than trying to govern every possible use of AI, we’ve focused on creating shared guidance with an intentional focus on ethical considerations, providing institutionally supported tools and helping people make responsible choices within the context of our educational mission.”
That same instinct shapes how Bowdoin approaches adoption on the ground. “It shows up in our ability to start with people and pedagogy rather than the technology,” she explains. “We can work directly with faculty through both our shared governance structures and personalized consultations to understand what they are trying to accomplish in a particular course, discipline or research context, and then explore where AI adds value, and where it may not.”
The result is a version of AI adoption that reads less like a technology rollout and more like an extension of what a liberal arts education is supposed to do in the first place: teach people to think critically and with a commitment to the common good.
Balancing Scale and Flexibility
The University of Hawaiʻi System faces a different kind of complexity. Ten campuses, a wide range of missions and a responsibility to ground AI strategy in something more specific than ‘best practices’.
“We are currently in a pivotal stage of ensuring that our AI strategy is built on a foundation that is authentically our own,” says Dr. Gloria Niles, Chief Academic Technology Innovation Officer for the UH System. A Hawaiian Culture and Values Task Force is drafting an AI Values Framework meant to guide decisions on everything from vendor selection to data governance to instructional support, rather than retrofitting existing, non-localized policies to fit.
Dr. Niles is candid about the tension that comes with operating at that scale. “We are constantly navigating the friction between the need for cohesive, high-level governance and the reality of diverse campus missions,” she says. “We need system-wide guardrails, especially on data sovereignty, security and privacy, because failing to align on these core infrastructure issues exposes the entire system to unnecessary risk. Yet if we default to a one-size-fits-all approach, we stifle the grassroots innovation essential to meaningfully integrating AI into teaching and learning.”
She points to a specific example of what that tension demands. UH’s Indigenous-serving mission carries a kuleana, or responsibility, to protect Hawaiian knowledge systems and intellectual property, which requires a centralized, culturally grounded stance on data governance. But the best AI practice in a professional-technical program at a community college looks nothing like the best practice in a humanities course at a research-intensive campus. Rather than force a single answer, UH has shifted its governance model from broad representative task forces toward structures built around subject-matter expertise, aiming to hold the line on shared risk while leaving room for campuses to work out what fits their own context.
“The tradeoff is that we must commit to a discursive process,” Dr. Niles says, “constantly negotiating alignment across campuses, rather than seeking a single, static policy that covers every scenario.”
Leading With Learning Before Legislation
At the University of Texas at San Antonio, the starting point wasn’t a values framework or a governance structure. It was an admission.
“When we first started planning our work around Generative AI, I felt like we didn’t know enough to even think about what policies would make sense,” says Dr. Melissa Vito, Vice Provost of Academic Innovation at UTSA. Existing policies, including those on academic integrity, seemed sufficient for the moment, and Dr. Vito was wary of writing something more permanent into an area that was moving too fast to pin down. “It was critical that we lead with curiosity and learning rather than fear and protection of what currently existed.”
So instead of policy, UTSA started with people. In January 2023, the university brought together more than 40 faculty members to understand what they were feeling and what they needed. That group helped shape a set of underlying principles rather than rules: that AI literacy would likely matter for the future workforce, that AI should amplify creativity rather than replace it and that critical thinking needed to stay central no matter how the curriculum used the technology.
That early faculty group has since grown into a peer learning network of more than 150 members spanning disciplines from math to astronomy to cybersecurity. “Faculty wanted to understand what was going on and how Generative AI could be used,” Dr. Vito says.
The group met regularly, brought in outside experts monthly and eventually developed the guidelines that UTSA still builds from today. By 2024, the model had grown beyond UTSA itself, with faculty and staff from across the UT System joining a shared convening to compare notes.
Context First, Framework Second
Three institutions. Three very different approaches. Yet each arrived at the same conclusion: AI strategy is strongest when it’s rooted in institutional mission rather than borrowed from someone else’s playbook.
Whether the challenge is preserving cultural knowledge, engaging a close-knit academic community, or helping faculty navigate emerging technologies, context shapes the questions leaders ask, the risks they prioritize and the opportunities they pursue.
In other words, the future of AI in higher education may not be defined by a single best practice, but by each institution’s ability to translate its values into action.
To hear how these leaders are turning institutional values into practical AI strategy, join Context Matters: Designing AI Strategy for Mission-Driven Institutions at EDUCAUSE on Thursday, October 1, 2026. Dr. Cristi Ford will moderate a conversation with Dr. Sherri Braxton Castanzo, Dr. Gloria Niles, and Dr. Melissa Vito as they share lessons learned, challenges encountered and insights for institutions charting their own AI journey.
Discover how to translate your institution’s values into AI action. Browse the Learning Lab for frameworks and strategies, then join us at Booth #1223 at EDUCAUSE for live demos and conversations with campus leaders.
Written by:
