The literature proposes several ways forward to ensure that AI tools support, rather than hinder, person-centred doctor-patient relationships. A handful of studies sug-gest that when AI is used as an assistive tool, this may have a positive impact on the doctor-patient relation-ship (e.g. Eysenbach et al.; Szalai). However, it is argued that patients and doctors may be unlikely to accept a shift to AI-led medical care, and such a shift could harm the doctor-patient relationship as AI tools are incapable of reproducing inherently human qualities of empathy and compassion. In the longer term, the debate is still open with regards to how human preferences for AI-led healthcare will evolve. Patients and doctor alike might start favouring the increased accuracy of AI-led care. However, current evidence regarding human preference points to the fact that this is not yet the case. There is broad agreement in the literature that the impact of AI on the doctor-patient relationship will influence and be influenced by the education of medical professionals. Most authors seem to suggest that medical education should focus on AI literacy and emotional intelligence, with some emphasising the importance of one over the