A Family Medicine Shared Vision and Road Map for AI in Primary Care

Introduction

Primary care is at a pivotal moment, facing systemic challenges while standing to benefit from emerging technologies. Despite its foundational role in improving outcomes and reducing costs, the field is under-resourced, overburdened and increasingly fragmented.^1,2 Amid these challenges, artificial intelligence (AI) presents an opportunity.³ In May 2025, the American Academy of Family Physicians (AAFP) and Rock Health brought together over 80 stakeholders at the Starfield Summit on Advancing AI and Digital Health in Primary Care to collaboratively create a vision of how AI can be harnessed to strengthen—but not supplant—the core values of family medicine and primary care. The summit reaffirmed the enduring relevance of Barbara Starfield’s 4 core attributes of primary care—first-contact access, continuity, comprehensiveness, and coordination—and emphasized the need for AI to support these principles. In this article, the family physicians on the summit’s planning committee summarize the shared vision, outline a road map for action, and reflect on the convening itself as a model for cross-sector collaboration. The road map includes action steps focused on stakeholder convening, advocacy for policy reform and resource development. It provides physician leaders, technology developers, and policy makers with a framework to guide responsible AI integration. A full report from the Starfield AI summit is available on aafp.org at https://www.aafp.org/dam/AAFP/documents/practice_management/ai-road-map.pdf.

A Shared Vision: AI That Strengthens Primary Care

The Starfield AI summit brought together over 80 stakeholders—including family physicians, technologists, payers, policy makers, and patient advocates—to co-create a vision for AI in primary care. The central thesis that emerged was clear. AI can strengthen primary care to improve human health, but only if:

• Primary care stakeholders are involved throughout the innovation life cycle

• Training and support are provided to clinicians and care teams

• Key systemic barriers are proactively addressed

Participants envisioned a potential future in which AI, when responsibly implemented, could serve as a trusted member of the care team, capable of triaging patients, summarizing charts, drafting educational materials, aiding clinical decision-making, and flagging care gaps. Emerging AI tools, when developed and implemented thoughtfully, could restore time for human connection, reduce administrative burden, and enhance the core attributes of primary care: first-contact access, continuity, comprehensiveness, and coordination.⁴

The Road Map: From Vision to Action

To translate this vision into reality, the summit produced a road map organized around 3 strategic pillars.

1. Convening the right stakeholders

   • ○
     Innovation networks: Establish national networks of primary care practices to pilot AI tools and inform real-world development (eg, family medicine AI centers of excellence)
   • ○
     Peer-buyer networks: Create transparent forums for practices to share information about experiences, costs and outcomes related to AI adoption
   • ○
     Governance standards: Develop primary care-specific benchmarks for safety, equity, and transparency of AI tools
2. Advocating for policy and payment reform

   • ○
     Payment models: Align reimbursement with AI-enabled care, including asynchronous preventive services
   • ○
     Legal frameworks: Clarify liability, oversight, and adverse-event reporting for AI use in clinical settings
3. Building resources and tools

   • ○
     Evaluation frameworks: Develop standardized frameworks that primary care practices can use to evaluate vendors and assess the return on investment of AI tools
   • ○
     AI literacy: Integrate AI education into undergraduate and graduate medical education, CME, and executive training
   • ○
     Patient engagement: Develop multilingual, accessible tools to support informed consent and promote trust in AI-supported care

The Convening: A Model for Cross-Sector Collaboration

The Starfield AI summit itself was a powerful demonstration of what is possible when diverse voices come together. Participants included frontline family physicians, internists, pediatricians, academic leaders, health system executives, electronic health record (EHR) vendors, and global AI experts. The summit’s structure—rooted in co-creation, not presentation—enabled candid dialogue and consensus building. It also surfaced tensions that exist between innovation and equity, speed and safety, and automation and humanism.

Importantly, those who attended the summit did not shy away from the hard questions. Participants debated the risks of AI exacerbating fragmentation, bias, and burnout. They acknowledged that AI alone cannot fix structural issues such as underfunding and workforce shortages. But they also affirmed that, if guided by the values of primary care, AI could help reverse decades of underinvestment and fragmentation.

Brief Reflections on Challenges and Core Attributes

The summit’s vision is grounded in a realistic understanding of primary care’s challenges: chronic underfunding, workforce attrition, rising complexity of care, and declining continuity. AI is not a panacea, but it offers tools to address some of these issues, particularly in workforce support, access, and care coordination.

Crucially, the summit reaffirmed the enduring relevance of Barbara Starfield’s 4 core attributes of primary care: first-contact access, continuity, comprehensiveness, and coordination. These must remain the North Star, guiding the standards by which all AI tools are judged. AI should not merely optimize workflows. It should deepen relationships, enhance trust, and improve outcomes. We have a real opportunity to transform primary care, empowering it to excel at its core attributes, and help the health care system achieve the Quintuple Aim.⁵

Conclusion

The future of primary care is being shaped now. The Starfield AI summit offers a hopeful, actionable vision for how AI can support—but not replace—the deeply human work of primary care. Realizing this vision will require shared responsibility, sustained collaboration, and a steadfast commitment to the values that define our field.

As physician leaders and policy makers, we have a choice. We can either let AI happen to primary care, or we can shape how it happens with intention, equity, and purpose. The “Starfield signal” is clear. Let’s answer it.