Use of AI in education risks a set of long-range and indirect impacts that may be difficult to assess. These include the risks of AI
1. Reifying assumptions, acting to label learners and teachers and thus shape their lifelong trajectories, rather than providing useful information to support agency in shaping those trajectories. Data in learning environments thus has potential to be misused outside those environments, including by employers.
1. Ossifying approaches, through 'locking in' particular models of learning and assessment via algorithms, structures of data capture and use, and data infrastructure that may be hard to change, particularly here they act to reduce skills or labour capacity to intervene. Learning environments must be dynamic to learners and their setting. The values that underpin technologies may not always be made clear up front, and evidence of their efficacy may be treated as a static given, rather than an evolving body of knowledge.
1. Standardising values, through provision of models built on single sets of static values that may not change across culture, context, or time. In this way the potential of AI to personalise content may be restricted to a narrowly individualistic view of personalisation and adaptivity.