Across Aotearoa New Zealand’s education sector, generative AI is rapidly becoming embedded in everyday practice. Much of the discussion understandably focuses on capability and governance. Yet as generative AI becomes integrated into the infrastructure of education, another question begins to emerge: What happens to institutional values when the systems supporting teaching, learning, and decision making begin participating directly in the practices through which educational reasoning occurs? In our recent article for The Conversation, my colleague Dr Rhiannon Lloyd and I explored this issue through the concept of value drift. The term describes how technologies can gradually influence how organisational principles are interpreted and enacted over time. Rather than replacing values outright, technologies can reshape how those values are expressed in everyday practice. This possibility is particularly significant in education, where institutional legitimacy depends heavily on trust in how decisions are made and justified. AI is becoming part of the infrastructure of education Educational systems have always been shaped by the technologies through which knowledge is organised and communicated. The textbook standardised curricular knowledge. Learning management systems structured the delivery of digital courses. Search engines transformed how students and educators access information. We have seen something similar in the case of social media which did not just test our existing ideas about privacy. It gradually changed them. What once felt intrusive or inappropriate now feels normal to many younger users. Generative AI represents another step in this evolution, but it operates differently from earlier tools. Rather than simply storing or retrieving information, these systems generate text that supports educational processes. AI systems can be used to produce feedback on student work, summarise complex material, generate explanations, draft course resources, and support institutional communication. As these capabilities become embedded in educational platforms, generative AI is beginning to shape the language through which teaching, learning, and evaluation take place. Where organisational values actually live When educational institutions articulate their values, they often point to formal statements. Universities emphasise academic integrity, intellectual independence, and fairness in assessment. Schools highlight care, equity, and the development of critical thinking. EdTech firms frequently emphasise access, learner success, and trust. Yet values rarely operate primarily at the level of policy statements. They are enacted through everyday professional judgement. They appear in how educators explain feedback to students, how institutions justify assessment decisions, and how digital platforms recommend learning pathways. In these moments, abstract principles such as fairness or transparency take concrete form. Once generative AI begins assisting in producing this language, it becomes part of the environment in which those interpretations take shape. How value drift emerges Value drift rarely occurs through dramatic ethical failures. Instead, it tends to emerge gradually through small shifts in practice. As educators and administrators incorporate AI tools into their workflows, certain patterns of explanation may become easier to produce than others. Feedback may be generated more quickly, explanations may become more standardised, and particular forms of reasoning may become more common because they are readily produced by AI systems. Over time, these shifts can accumulate. Institutions may continue to affirm the same principles, yet the way those principles are enacted in everyday practice can evolve. What counts as “good feedback,” a “fair explanation,” or a “transparent justification” may subtly change as AI-mediated communication becomes routine. Recognising this possibility does not imply that AI adoption is problematic. Educational technologies have always shaped institutional practice. However, generative AI operates in the communicative space where educational reasoning occurs, making these shifts particularly important to notice. What this means for educators, EdTech firms, and policy makers For educators, the concept of value drift highlights the importance of reflecting on how AI-assisted practices shape teaching and assessment. Tools that generate feedback, explanations, or learning resources can offer substantial benefits, particularly in large-scale teaching environments. At the same time, educators may need to remain attentive to how these tools influence the character of feedback, the dialogue with students, and the reasoning behind evaluation and support. For EdTech firms, the implications extend to product design. As generative AI features become central to educational platforms, design decisions inevitably influence how teaching and learning practices unfold. The ways AI systems structure feedback, generate explanations, or recommend learning pathways can shape how educational values are enacted within institutions. Designing these systems therefore involves not only technical considerations but also questions about how professional judgement and educational reasoning are mediated through technology. For policy makers and sector leaders, the discussion suggests that responsible AI governance must go beyond establishing high-level principles. Frameworks emphasising fairness, transparency, and accountability remain essential. However, the concept of value drift highlights that values are continually interpreted through practice. Effective governance may therefore require ongoing attention to how AI-enabled systems influence everyday institutional reasoning across the education sector. A capability the sector will increasingly need As generative AI becomes more deeply embedded in educational systems, the challenge for the EdTech ecosystem will not simply be to adopt the technology responsibly. It will also be to remain attentive to how the presence of these systems gradually reshapes educational practice. Technological change rarely transforms institutions overnight. More often it alters habits, language, and expectations incrementally until a new normal quietly emerges. For educators, technology developers, and policy leaders across Aotearoa, recognising these gradual shifts may become an important capability in ensuring that the values underpinning education remain actively interpreted rather than passively reshaped by the tools we build and use. The post When AI quietly shifts what organisations value appeared first on Tech New Zealand.
