Ulster University study: Students normalised AI by 2026, staff wary
A 1,665-person longitudinal study reveals a widening student-staff gap on AI use.
A longitudinal study by Gerard et al. at Ulster University (arXiv:2607.16223) tracked AI perceptions across three survey waves from 2024 to 2026, surveying 1,665 undergraduates, doctoral researchers, teaching staff, and non-teaching staff. The results paint a stark picture: students have rapidly normalised generative AI tools like ChatGPT, moving from tentative experimentation to routine, integrated use. Meanwhile, teaching staff express persistent concerns about academic integrity, assessment validity, and the erosion of critical thinking skills.
Doctoral researchers and non-teaching staff occupied an intermediate position, balancing pragmatic adoption with institutional caution. The gap between student behavior and staff perceptions widened over the two-year period as institutional policies struggled to keep pace with real-world usage. The study moves beyond single-point snapshots common in prior research, demonstrating how attitudes evolve in real time.
The findings carry critical implications: universities need adaptive policies that acknowledge the reality of AI use, robust AI literacy initiatives for all stakeholders, and targeted training for staff to address assessment redesign and integrity challenges. Without proactive measures, the disconnect between practice and policy will only deepen.
- 1,665 participants across three waves (2024-2026) at Ulster University.
- Students shifted from tentative experimentation to routine AI use; staff remained concerned about integrity and critical thinking.
- Student-staff gap widened as institutional policy failed to keep pace with AI adoption.
Why It Matters
Universities must adapt policies and training quickly or risk a growing disconnect between actual AI use and academic integrity.