Professor Edward Harcourt is Director of Oxford's Institute for Ethics in AI, based in the university's philosophy department. His work covers AI ethics, AI safety and AI governance, including questions of regulation, liability and where to strike the balance between oversight and innovation, but his particular focus is on what philosophy has to say about human mindedness now that AI has made those questions urgent rather than abstract.

Edward argues that human beings possess something not reproducible by any artificial mind, however cognitively sophisticated: an embodied existence that comes bundled with mortality, appetite, suffering and the capacity for touch, all of which shape human relationships in ways a chatbot cannot replicate. He is careful to distinguish this from a simple safety argument. The real risk of chatbots, in his view, is not that they reshape social life but that people are led to believe there is something real on the other end of the conversation.

He identifies two further risks that fall outside conventional safety debates: cognitive atrophy, as people lose the habit of working things out for themselves, and social atrophy, as convenience quietly displaces the ordinary human contact that people need in order to thrive.