AI undermining degree value & critical AI literacy
Key Questions
What AI policy has the University of Manchester adopted?
The University of Manchester has implemented a four-tier AI policy for student assessments to guide appropriate use across different contexts. This approach aims to balance innovation with academic integrity amid rising AI adoption.
How do AI detectors impact international students according to HEPI?
HEPI warns that seven widely used AI detectors misclassified over 61% of essays written by non-native English speakers, leading to potential unfair penalties. This has prompted calls for assessment redesign and policy reform to avoid harm.
What does the OECD report reveal about GenAI governance in UK higher education?
The OECD reports that 95% of UK undergraduates use GenAI, yet only 38% of institutions provide secure tools and 19% have formal policies. It urges stronger, systematic governance to address the gap between adoption and oversight.
What strategies are recommended for integrating AI in higher education?
Research highlights a shift from avoidance to adoption, with discipline-specific approaches and faculty development models like the 'buffet' framework to support evolving AI use. These emphasize staff training and student literacy for inclusive integration.
How is AI affecting lifelong learning and inclusion in higher education?
AI is reshaping lifelong learning, academic judgment, and equity, but it does not automatically benefit all learners. Experts call for targeted staff development and student AI literacy to promote inclusive outcomes.
What is Ofqual's regulatory approach to AI in qualifications?
Ofqual focuses on fairness, validity, and security in its formal regulatory framework for AI use in qualifications. This provides a key signal for higher education institutions developing assessment policies.
Why are students protesting AI-taught classes in UK higher education?
Students have protested an AI-taught programming apprenticeship, citing hypocrisy and a trust gap, arguing they could have used ChatGPT themselves. This underscores broader concerns over AI's role in teaching and learning.
What practical strategies exist for designing AI-resilient assessments?
New resources provide strategies for UK HE institutions to redesign assessments that mitigate AI misuse while maintaining integrity. These include three-band frameworks distinguishing permitted, grey, and prohibited AI use in coursework and theses.
LSE Impact piece argues AI can game rubrics. BSA survey: 34% think degrees not worthwhile. Calls for critical AI literacy. New frameworks (VALID, AI Assessment Scale) and case studies. HEPI Policy Note, QAA response, Leicester call. Global survey: 15% see AI integrated, 29% trust faculty, 60% worry about peer misuse. Manchester study on assessment redesign. HEPI induction piece. Assistive technology equity concerns. A new systematic review (33 studies) on AI for academic dishonesty detection/prevention finds AI should complement human judgment, not replace it. A new academic reply argues AI literacy insufficient. A new Turnitin-based cross-country comparison shows US student AI writing rates double UK levels (9.8% UK vs ~20% US), highlighting detection vs usage gap. The AI Adoption Surges report adds: only 32% have central AI policy, assessment redesign split, cheating concerns rising. A new HEPI piece explores how AI may be eroding the social core of student teamwork. A new practical article debunks the 'percentage myth' in AI academic writing, mapping a three-band framework (permitted, grey, prohibited) and distinguishing rules for coursework, theses, and journals. A new QAA report confirms rampant policy variability causing student confusion and trust erosion, with powerful student quotes; calls for a community of practice. A new THE study finds most students deny using AI when forbidden, with 32% 'conscientious objectors' and international students more reliant on AI – highlighting the need for co-created policy. A new Edinburgh Napier study (6,000+ respondents) finds 67% claim honesty, 5.2% habitual rule-breakers, 32% conscientious objectors, international students heaviest users, reinforcing the trust gap and policy inconsistency. Bristol's AI principles offer a practical institutional model. A new LSE article explores designing AI tutors that challenge rather than pander, offering practical design principles for UK HE. Ofqual has published its formal regulatory approach to AI in qualifications, focusing on fairness, validity, and security – a key signal for HE assessment policies. A new student protest over an AI-taught UK programming apprenticeship highlights hypocrisy and trust gap. ICS.AI free AI offer could level access but raises vendor lock-in concerns. A new PDF on designing AI-resilient assessment provides practical strategies for UK HE, directly addressing assessment redesign and academic integrity. New OECD report (2026) adds: 95% of UK undergraduates use GenAI, but only 38% of institutions provide secure tools, and only 19% have formal policies – governance lagging behind adoption. A new 'buffet' model of faculty development offers practical strategies for supporting diverse, evolving AI use across disciplines. A new HEPI piece warns AI detectors misflag international students (61% misclassification of non-native writing), with OIA cases showing real harm, strengthening the case for assessment redesign and policy reform. A research article from Northern Ireland provides evidence-based discipline-specific strategies for AI integration. An article on AI, lifelong learning and inclusive higher education challenges the assumption that AI automatically benefits all learners, calling for staff development and student AI literacy.