BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
 M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20261120T145613Z
DTSTART:20261124T120000Z
SUMMARY:Maths education seminar: Igor Chernyavsky on GenAI in STACK e-ass
 essments
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}vjv-mwp3qh0
 7-d1mq5z
DESCRIPTION:Speaker: Igor Chernyavsky (University of Manchester)\n\nTitle
 : GenAI in STACK e-assessment: a friend or a foe?\n\nAbstract: Rapid adv
 ances in generative AI (GenAI) are reshaping electronic assessment\, inc
 luding assessments delivered through STACK\, a computer-algebra-based sy
 stem for Moodle. This talk will explore the potential of this technology
  to empower teachers and students\, as well as the challenges associated
  with the integrity of non-invigilated assessments. First\, we explore t
 he still poorly understood question of how students engage with GenAI du
 ring assessments. We present a case study of two online time-restricted 
 tests in advanced engineering mathematics delivered through STACK. We de
 ployed ‘AI friction’ techniques: individualised watermarking and hidden 
 prompt injection to detect and quantify GenAI use in low-stakes tests. T
 hese approaches were robust across the GenAI models tested and provided 
 limited evidence of compromised academic integrity across a large cohort
 . We discuss how AI friction could incentivise self-study in future form
 ative and summative assessments. Finally\, we demonstrate the potential 
 of GenAI coding tools to support the design and quality assurance of STA
 CK quizzes. This work has been done jointly with Colin Steele and Anthon
 y Thornton.
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:Frank Adams 1 (and online)\, Alan Turing Building\, Manchester
END:VEVENT
END:VCALENDAR
