As patient-facing medical AI tools—particularly chatbots—become increasingly integrated into healthcare, they raise important ethical challenges that go beyond technical accuracy or safety. This chapter explores some of these key challenges posed by these particular AIs through the lens of relational ontology and narrative identity. While these tools offer remarkable diagnostic capabilities and promise improved access and efficiency across healthcare systems, they risk undermining the relational structures that constitute personal identity. The chapter argues that patient-facing medical AIs, by mimicking relational cues without true recognition or presence, may lead to ontological confusion, relational isolation, and the displacement of moral responsibility. Patients may be ‘seen’ through a clinical gaze that flattens their complexity into optimised, legible data points, and thereby erodes communal health narratives and dialogical self-understanding. These risks are not peripheral but structurally embedded in systems incentivised to replace the moral labour of care with algorithmic compliance. This chapter concludes by affirming that while medical AI can support diagnosis and treatment planning, it must never replace the human moral agent. Instead, AI must be built to complement, not displace, the relational work that is essential to creating and sustaining personal identity through care.