Presentation Schedule
Risk Framing and Emotional Tone in University AI Policies: A Computational Discourse Analysis of Global Higher Education Governance (107552)
This presentation will be live-streamed via Zoom (Online Access)
Monday, 13 July 2026 12:05
Session: Session 2
Room: Live-Stream Room 3
Presentation Type:Live-Stream Presentation
– click here to convert to your timezone
Purpose: This study examines how top-ranked universities globally frame artificial intelligence governance through linguistic analysis of risk intensity and emotional tone in official AI policy documents. In the context of AI governance, where stakeholders experience emotions ranging from enthusiasm to apprehension (Güneş & Kaban, 2025; Kim, 2025), the emotional register of policy language becomes a critical variable in governance effectiveness.
Design/Methodology: We conducted quantitative text analysis using LIWC-22 software on AI policies from 20 top-ranked universities (Times Higher Education 2026 rankings) across four geographic regions. The analysis examined emotional tone, anxiety language, risk framing, and affective balance across 54,093 words of policy text. Findings: Universities exhibit substantial variation in risk framing intensity (3.4-fold difference) driven primarily by tentative language rather than explicit anxiety. Eighty percent of policies adopt a negative/uncertain emotional tone (M=43.06, SD=8.70), with significant regional differences: UK universities demonstrate confident optimism (M=49.66), US universities show moderate uncertainty (M=44.26), and Asian universities express pronounced caution (M=32.75, p=0.072). Asian institutions use significantly more anxiety-related language (p=0.023) and maintain the most negative affect balance (p=0.043), despite comparable overall risk intensity across regions (p=0.728). Originality/Value: This study contributes a systematic, computational examination of emotional framing within university AI governance policies—an area that has seen limited formal investigation to date. The findings reveal that effective AI governance communication is culturally situated, challenging assumptions about universal policy language and demonstrating that institutions reach similar risk conclusions through different emotional pathways.
Authors:
Stanley Terkuma Asongo, University of Massachusetts Lowell, United States
About the Presenter(s)
Ph.D. student at University of Massachusetts Lowell
Connect on Linkedin
https://www.linkedin.com/in/terkuma-asongo-050b47261
See this presentation on the full schedule – Monday Schedule
Conference Comments & Feedback
Place a comment using your LinkedIn profileShare this Presentation
Comments
Powered by WP LinkPress