UNIVERSITY OF ATHENS POSTDOCTORAL RESEARCHER FEATURES IN AI HUB FOR EDUCATION RESEARCH STUDY REPOSITORY (SCALE INITIATIVE, STANFORD ACCELERATOR FOR LEARNING)

A paper by postdoctoral researcher at the Department of Pedagogy and Primary Education, Tryfon Sivenas, working under Professor Emeritus Constantine Skordoulis, was included in the AI Hub for Education Research Study Repository of Stanford's SCALE Initiative (January 2026 update). It was one of three studies explicitly highlighted (the other two being from UCL and Carnegie Mellon) with direct reference to the National and Kapodistrian University of Athens.

The paper also features in three Stanford LinkedIn posts:

https://www.linkedin.com/posts/chris-agnew-74bb8217_its-been-a-couple-months-since-youve-activity-7419903005375950848-MBpI
https://www.linkedin.com/posts/ai-education-stanford_in-this-first-edition-of-the-new-year-we-activity-7424538110652510208-Oa9d
https://www.linkedin.com/pulse/anticipating-different-ai-futures-schools-according-s3w9c

Titled 'ChatGPT-5 in Secondary Education: A Mixed-Methods Analysis of Student Attitudes, AI Anxiety, and Hallucination-Aware Use', Tryfon Sivenas's research is among the first internationally to examine how secondary students respond to AI hallucinations - that is, the generation of false or misleading content by AI systems. Drawing on a study of 109 students across three Greek high schools, Sivenas introduces the concept of 'epistemic safeguarding': after encountering hallucinations, many students reported limiting AI use to topics where they already had knowledge and could verify the accuracy of answers.

The full study is available in the Stanford Repository and as a preprint in arXiv.