The rapid uptake of generative AI in education has intensified longstanding questions about wellbeing, agency, and responsibility for both teachers and learners. While institutional responses often focus on regulation, skills development, or efficiency gains, this concise paper argues that wellbeing in an AI mediated world is fundamentally a learning design issue. From a postdigital perspective, wellbeing cannot be reduced to individual resilience or coping strategies; it emerges through more than human entanglements involving people, technologies, emotions, institutional cultures, and material conditions.

Postdigital scholarship highlights that learning and teaching now occur within hybrid assemblages of humans and non-humans, where agency is distributed and outcomes are uncertain. In these contexts, affect is not peripheral. Teachers and learners encounter AI with mixtures of curiosity, excitement, anxiety, moral concern, frustration, and uncertainty. These emotional responses shape judgement, critical thinking, willingness to experiment, and perceptions of responsibility. Treating emotions as distractions to be managed risks undermining both learning and wellbeing.

For learners, AI promises personalisation and support, yet can also generate pressure, surveillance anxiety, and confusion around authorship and ethical boundaries. For teachers, AI introduces affective labour: negotiating institutional expectations, modelling ethical practice, and supporting students emotionally while managing their own uncertainty. Wellbeing, therefore, cannot be addressed without considering how learning designs invite—or suppress—emotional engagement with AI.

This paper proposes that learning designs explicitly engaging the affective domain are central to wellbeing in an AI world. Such designs move beyond individualised performance metrics towards relational, reflective, and embodied practices. Indigenising perspectives, particularly from Aotearoa New Zealand, offer valuable insights by foregrounding care (manaakitanga), relationships (whanaungatanga), and holistic conceptions of learning that align closely with postdigital thinking. These approaches challenge neoliberal assumptions that value only what is measurable or automatable.

Embodied and materially focused activities—such as reflective engagement with AI outputs, collective sense making, or situated encounters with data—can support both learner and teacher wellbeing by legitimising emotion, uncertainty, and dialogue. Rather than seeking control or certainty, such designs acknowledge that thriving in an AI world involves learning to feel, judge, and relate well within more than human contexts.

Discussion Points for Conference Participants

• How might learning design support emotional awareness rather than suppressing affect in AI mediated education?

• What responsibilities do institutions have for teacher wellbeing when AI intensifies affective labour?

• How can indigenous and relational approaches inform wellbeing focused AI pedagogy across contexts?