From fear to innovation: A case study of transformative faculty development for ethical AI integration in higher education

Synthesis: Chick, Morello and Staffey document a six-week faculty institute at the University of Bridgeport in Summer 2025 that took ten instructors and staff from Education, Health Sciences, Business, Sport Management, Traditional Chinese Medicine and Social Sciences through Universal Design for Learning, Bloom's Taxonomy and design thinking as lenses for generative AI integration. Surveys and reflections and ten capstone redesigns show a cohort that entered with low AI fluency and left with 60% reporting high confidence in guiding student AI use and 90% reporting positive perceptions. The authors name four themes, fear shifting to curiosity, a desire for ethical clarity, inclusive design as an equity amplifier, and faculty moving from gatekeepers to guides, and propose "symbiotic pedagogy" as the framework their participants actually built. Their conclusion is that resistance was never the problem: the barriers surviving the institute are institutional, a policy vacuum, unequal tool access, unrecognized time, and Assessment systems built for pre-AI work.

Key Findings

1. A deliberately mixed, very small cohort. Ten faculty completed UBRIDGE, six weekly two-hour sessions; teaching experience ran from 2 to 35 years and recruitment was purposive rather than enthusiast-only.
2. Confidence and perceptions moved sharply in one direction. Initially 80% rated their confidence in guiding student AI use as low or nonexistent and 20% held negative views of AI in higher education; afterwards no one reported no confidence, 60% placed themselves at the top of the scale, and negative views fell to zero.
3. Concerns centered on integrity and over-reliance, not replacement. Academic integrity led pre-survey concerns (80%), then over-reliance (70%) and bias or inaccuracy (50%); equity and access were named by none.
4. The capstone redesigns are the hardest evidence. All ten redesigned a course component: 8 of 10 positioned AI as a scaffold rather than a replacement, 7 of 10 redefined their teaching purpose toward inquiry, and 6 of 10 embedded critique of AI bias.
5. Intentions were near universal; the enabling conditions were not. All ten said they would keep integrating AI and recommend the institute, yet the same participants described contradictory policy signals, personal subscriptions for continued tool access, and no time support for the redesigns they had planned.

What the institute was, and how it was studied

The design refused to treat AI as tool training: UDL as the inclusion frame, Bloom's Taxonomy to locate where AI helps and where it displaces a learning objective, design thinking as the route from theory to course design, plus Technological Pedagogical Content Knowledge (TPACK) as the basis for the authors' "AIPACK" and Mezirow's transformative learning. Methodologically it is a qualitative case study with survey components, framed as pragmatic constructivism and analyzed with Braun and Clarke's thematic analysis. Evidence came from pre and post surveys, design worksheets, asynchronous reflections, capstone submissions and facilitator field notes. The lead authors were themselves the facilitators and kept a reflexive journal about that dual role.

What changed over six weeks

The vocabulary shift is the clearest marker: early descriptions of AI as a "cheating machine" or a "dehumanizing force" gave way to "creative partner" and "curious assistant". A mathematics professor's early question — "Now a machine can solve problems faster and explain solutions better than I can. What's my value anymore?" — captures the fear the institute had to work through, and the authors read the confidence jump as a reorientation of professional identity, from reacting to AI to designing for it.

Four themes, and the framework the authors propose

Fear moved to curiosity across four themes: ethical clarity wanted in principled form rather than rule lists, producing AI collaboration statements and a decision tree weighing learning objectives, assessment validity and student agency; inclusive design as an equity amplifier, including an experiment whose stereotyped career advice became a teaching case about bias; and faculty moving from gatekeepers to guides, with assignments shifting from "AI-proof" to documented, defensible collaboration. Their theoretical contribution is "symbiotic pedagogy": complementary design, transparent process, critical integration, embedded ethical reasoning, and iterative refinement that treats experimentation and failure as the route to competence. For institutions they add a five-domain support framework and warn that piecemeal responses leaving policy and infrastructure untouched will fail even enthusiastic cohorts.

What this means for practice

• Instructors. Open with the professional-identity question rather than tool demos; that is what moved this cohort. Test each AI use against the learning objective it serves, and reuse the paper's own bias exercise — one prompt, different demographic markers, critique the difference.
• Faculty developers. Six weekly two-hour sessions with a required capstone redesign moved a purposive, not enthusiast-only, cohort. Stipends and individual facilitator feedback were part of the treatment, so budget both.
• Administrators. The barriers that survived were structural, not attitudinal: contradictory policy signals, personally funded tool access, unrecognized redesign time and Assessment built for pre-AI work. Unless promotion criteria recognize pedagogical scholarship alongside the paper's five support domains, individual enthusiasm is wasted.
• Researchers. The design is replicable: pre and post surveys plus artifact analysis of the redesigns, with a reflexive journal covering the dual facilitator role.

Limitations

• Ten purposively recruited volunteers at one institution, each receiving a stipend and individual facilitator feedback, so the transformation may not transfer to faculty required rather than invited to engage.
• Outcomes rest on self-report and facilitator-observed evidence, and the lead authors were themselves the facilitators, so participant narratives may have been shaped by that dual role.
• Six weeks can change what faculty believe and design, but not whether practice becomes habitual, and students were never consulted even in the design thinking empathy phase, so the redesigns encode faculty assumptions about student needs.
• The closing claim is diagnostic rather than instructional: resistance "stems not from technophobia or stubborn traditionalism but from legitimate concerns about educational quality, equity, and human agency". The conditions that let the work continue — clear ethical guidance, real time, institutional tool access and assessment that can handle human-AI collaboration — are not deliverables of a training program.

Citation

Chick, J., Morello, L., & Staffey, S. (2026). From fear to innovation: A case study of transformative faculty development for ethical AI integration in higher education. International Journal for Educational Integrity, 22(18).