Can AI Make Learning Less Stressful?

The educational landscape experiences a major shift in how students use generative AI (GenAI) tools. While in 2024, we saw ‘generating ideas’ was the most common reason for using GenAI, recent Harvard Business Review study’s (2025) findings indicate that ‘therapy/companionship’ is the leading use case, followed by ‘organising my life’ and ‘finding purpose’. This change indicates that students are no longer looking for academic assistance; they are turning to GenAI for emotional support and existential questions, fundamentally influencing the role of technology in their personal lives and educational journey. This flash talk covered the implications of this change through the lens of AIEOU’s four foundational pillars: Design, Regulation, Implementation, and Impact. Initial research findings indicate a critical paradox: while GenAI use in education demonstrates enhancement of short-term task performance, it inhibits students’ intrinsic motivation and deep learning transfer. Even more concerning is the emergence of ‘metacognitive laziness’, where unsupervised GenAI use can create overdependence on AI-generated responses, resulting in shortcuts, potentially leading to slower critical thinking development among students (Fan et al., 2025).