The second paper in the series takes the emergence of AI as its starting point, arguing that it both exposes and accelerates the inadequacy of the current model. The authors examine the implications of narrow AI, general AI and large language models for pedagogy and school organisation — from personalised curricula and administrative efficiency to risks of cognitive passivity, embedded bias and commercial capture of education. They argue that if any teacher who could be replaced by AI should be, then schools urgently need to articulate and defend what only human educators can do. The paper concludes that AI does not resolve the question of education’s purpose — it makes answering it more urgent.