AI could turn concierge medicine into the future interface to healthcare

Concierge medicine, despite its high cost, faces a significant limitation: physicians simply cannot provide sufficient attention to each patient due to the fixed number of working hours. Artificial intelligence (AI) could potentially alleviate this constraint by automating routine tasks such as gathering information, explaining test results, monitoring progress, and checking patient adherence.

This could allow concierge-style care to become more widely available, moving from a luxury service to a more accessible form of healthcare.

Several healthcare companies are already exploring this concept, developing systems that combine AI with clinician oversight. Hims & Hers, for example, plans to use AI to maintain treatment context between appointments and bring in clinicians when human judgment is required. Teladoc One merges medical history with claims, pharmacy, device, and medical-record data, supported by an always-on AI layer.

Amazon Health AI can interpret records, explain lab results, and route patients to One Medical for professional care when necessary.

The key difference lies not just in AI's ability to make doctors more productive, but in the potential to separate medical attention from physician minutes. AI can handle routine interpretation, follow-up, and coordination, allowing clinicians to support a far greater number of patient interactions. This shift could enable healthcare delivery to move from a snapshot-based system to one where continuous data allows for ongoing monitoring and timely interventions.

However, there are still significant clinical challenges to address, particularly with the accuracy of wearables and the ability of AI systems to distinguish meaningful changes from normal biological fluctuations. As more administrative and coordination work moves into software, the minimum economic size of a healthcare provider could decrease, potentially leading to more specialized services.

At the same time, AI can become more useful when it sees the entire patient context, including medications, diagnoses, lab results, symptoms, device measurements, previous treatments, and outcomes. This creates a push for care provision to fragment while patient context consolidates, potentially leading to a healthcare model where patients interact primarily with one health intelligence layer that understands their combined records. The healthcare company itself may not be the natural unit of integration.

While the company with the best medical AI may initially have an advantage, the economics of AI could make access to good general medical reasoning increasingly accessible, potentially reducing the defensibility of any single provider. The true advantage may lie in the ability of healthcare platforms to accumulate valuable longitudinal outcome histories and trusted clinical workflows, which could become proprietary assets.