The AI Doctor Is Coming — But Who Gets to Be in Charge?

It’s already woven into everyday life, almost imperceptibly, as people find themselves typing symptoms into AI before even considering a GP, not necessarily out of deeper trust, but because it offers something immediate, contained, and, in many ways, less demanding than trying to translate a complex internal experience into something legible within the constraints of a short appointment. There is something revealing in that instinct, in the quiet preference for a space that does not rush, interrupt, or require performance, particularly when the body itself feels unclear, contradictory, or difficult to articulate, and I recognise that pull, not as a rejection of care, but as a search for a different kind of encounter with it. Seen in this light, the idea of an AI doctor does not feel abrupt, but like an extension of an existing behaviour, one that is gradually reshaping how people navigate uncertainty, interpret their bodies, and locate authority in moments of vulnerability. Martin Varsavsky’s proposal sits precisely within that shift, not as a radical departure, but as an attempt to formalise what is already taking place by moving from interpretation into intervention, from a system that offers possibilities to one that is permitted to direct outcomes, prescribe treatments, and carry the legitimacy traditionally reserved for human practitioners. This movement from suggestion into action might appear procedural, yet it introduces a more profound transformation, because the capacity to intervene carries weight beyond efficiency, entering a space historically shaped by judgement, accountability, and the relational dynamics that have long defined medical practice. There is, undeniably, a sense of possibility here, particularly for those who have experienced dismissal, delay, or inconsistency within existing healthcare systems, where being met quickly, listened to without friction, and given clarity can feel not just convenient, but restorative. Access, in that sense, becomes more expansive. Yet access alone does not constitute empowerment, especially when the structure through which that access is delivered begins to alter how decisions are received, interpreted, and internalised. When a human practitioner offers guidance, it exists within a shared space that allows for questioning, hesitation, and negotiation, where understanding emerges through interaction rather than arriving as a fixed conclusion, and where the patient remains visibly present within the process of meaning-making. As this dynamic shifts towards AI, that interpretive space becomes less tangible, replaced by outputs that carry a tone of certainty, even when they are constructed from probabilities and patterns that remain largely unseen, making it more difficult to locate one’s own perspective within the exchange. This does not render the technology inherently problematic, but it does change the conditions under which trust is formed. At the same time, it would be incomplete to frame this solely as a loss, given that existing systems have already failed to meet many people with the depth, consistency, or attentiveness they require, leaving gaps that technologies like this are now attempting to fill. For some, these systems may feel more responsive, less intimidating, or better aligned with their need for clarity and autonomy. Which is precisely why the conversation cannot rest on acceptance or rejection alone, but must extend towards examining how such systems are shaped, whose interests they reflect, and how individuals remain active within their use. Because there is a version of this development that supports a deeper connection to one’s own body, offering insight without diminishing intuition, and expanding access without narrowing agency. There is also a version that moves in the opposite direction, where authority becomes increasingly diffuse, decisions feel less open to interpretation, and individuals begin to defer to outputs that appear more certain than their own internal sense-making. What emerges will depend less on the existence of the technology and more on the frameworks, values, and intentions that guide its integration. Redefining what constitutes a doctor, then, is not simply a technical adjustment, but a cultural shift, one that has the potential either to distance people further from their own embodied understanding or to support a more informed, engaged, and self-aware relationship with it. The distinction will not be made by the system alone, but by how consciously it is shaped, and by whether those using it remain present enough to question, interpret, and ultimately retain a sense of authorship over their own care.