Weekly AI News Update (21-27 November 2025)

CUHK’s new AI framework introduces a four-level system to guide faculty on AI usage in assignments, aiming to balance innovation with academic integrity. Vice-President Jin Guoqing emphasizes the importance of critical thinking and ethical AI use in education. 🔗 Ming Pao
Tsinghua University has issued institution-wide principles governing educational AI use, defining ethics, data security and academic integrity, restricting misconduct, and guiding teaching and postgraduate research within a “proactive yet prudent” framework. 🔗 Tsinghua University
Over 500 educators from 38 countries discussed ethical, human-centred AI integration. UNESCO highlighted teacher primacy, equity, critical evaluation, intellectual property, bias mitigation, and the 193-state Ethics of AI Recommendation guiding national education policy frameworks. 🔗 United Nations Educational, Scientific and Cultural Organization
Australian TALIS findings show widespread classroom AI use, especially among younger teachers, alongside benefits for planning and personalisation, and concerns over integrity, bias, reliability, and insufficient teacher skills. 🔗 Teacher Magazine
Self-supervised learning enables AI to derive training signals from unlabeled data, using pretext and contrastive tasks to build transferable representations, advancing applications in computer vision, language processing, speech recognition, and healthcare. 🔗 Netguru