This seminar will consider why—from a legal and ethical standpoint—space must be made for patient decision-making in the design of AI systems in healthcare. Implementing artificial intelligence in medicine is complicated by the web of relationships at play—and not least by the need to respect a patient’s human rights. Two talks from HeLEX researchers will address this issue from socio-legal perspectives. Miranda Mourby argues that some general responses to AI risk (e.g. novel insurance models) do not address the need for patient-centred dialogue following Montgomery v Lanarkshire Health Board. Minimum standards of transparency are thus also needed for medical AI under healthcare law; mirroring data protection requirements & the newly proposed EU AI Regulation. Nisha Shah then advocates for more and better engagement with patients in AI design and implementation, and the growing calls to tackle socio-technical barriers to involvement in medical decisions.