Establishing a National Framework for Family Medicine AI Centers of Excellence

Artificial intelligence (AI) is reshaping health care at a breakneck pace. From reducing administrative burdens to enabling next-generation clinical decision support, AI is already transforming how care is delivered, coordinated, and experienced. Family medicine (FM) has been part of this transformation from the start, shaping how AI can advance equity, access, and whole-person care.

Over the past 4 years, the family of FM organizations has laid essential groundwork to build AI capacity across the discipline. In 2021, the American Board of Family Medicine (ABFM) and the Center for Professionalism and Value in Health Care convened an AI summit that brought together early thought leaders to explore opportunities and challenges related to AI in primary care. This catalyzed in 2022 the ABFM Foundation’s “Enterprise AI and Building Long-Term (EnAIBL) Capacity for FM” initiative, a national collaborative supporting FM departments in strengthening the people, infrastructure, and processes needed to harness AI’s potential.

In 2023, through an effort funded by the Gordon and Betty Moore Foundation and facilitated by the ABFM, an AI Bootcamp series was launched at the North American Primary Care Research Group (NAPCRG) annual meeting. This was followed in 2024 by the development and dissemination of the Society of Teachers of Family Medicine’s (STFM) AI and Machine Learning for Primary Care (AiM-PC) curriculum. Additional momentum came from the American Academy of Family Physicians (AAFP) and Rock Health AI Starfield Summit in May 2025, culminating in a Starfield report ¹ that calls for a formalized “primary care innovation network” for AI.

Aligned with this call, STFM’s AI Task Force, in collaboration with the Association of Departments of Family Medicine (ADFM) and with funding from the ABFM Foundation, is leading a multiyear initiative to establish a national framework for Family Medicine AI Centers of Excellence (CoE). The goal is to help organizations build, sustain, and integrate AI capacity across clinical care, education, and research, not as separate domains, but as integrated capabilities that reflect the breadth and impact of our discipline.

The first phase of this initiative has focused on defining what excellence in AI looks like within FM. To that end, the STFM AI Task Force developed a draft CoE framework grounded in the following foundations:

1. The five Cs of primary care: first contact, comprehensiveness, continuity, coordination, and community;

2. The quintuple aim: patient experience, population health, cost of care, care team well-being, and health equity;

3. The three pillars of academic FM: clinical care, education, and research;

4. Leading models from existing health care and industry CoEs; and

5. Insights from FM leaders across the family of FM organizations working at the intersection of AI, innovation, and care delivery.

The initial development process was iterative and collaborative, shaped by landscape scans, feedback cycles, and structured working sessions. Every element was included based on its relevance, feasibility, and alignment with FM’s mission. The resulting draft framework is organized into four concentric domains, as shown in Figure 1, and outlined as follows:

• Purpose: a unifying mission to advance the Quintuple Aim through responsible use of AI.

• Core functions: clinical transformation, education and workforce development, research and evaluation.

• Enabling conditions: infrastructure, process, people, and culture.

• Illustrative activities: governance, funding, talent development, IT optimization, data access, community and provider engagement, partnerships, and more.

The framework is not intended to be prescriptive. Instead, it is the starting point for a practical and adaptable guide, built by and for the FM community, to help institutions assess their current capacity, identify gaps, and chart a path toward AI excellence that fits their mission.

By design, it emphasizes integration. AI in clinical practice must be connected to research and scholarship. Educational efforts must prepare learners to not only use AI, but to question and lead it. Infrastructure must support thoughtful implementation, ensuring tools reflect FM’s core values of equity and relationship-centered care. By anchoring the work in shared values and clear functions, we can support innovation while maintaining a consistent bar for quality and responsibility.

To ensure broad impact, the framework is being designed with all of FM in mind, especially those in rural communities and in safety net organizations. These settings are central to primary care and often face distinct barriers to AI adoption. The CoE initiative aims to bridge these gaps by fostering mentorship and innovation, sharing resources, and ensuring community-rooted care remains at the heart of AI advancement in FM.

The work ahead will unfold across several key phases:

• With the publication of this article, we're inviting feedback² from the FM community on the draft framework. This early input will shape future iterations and ensure it reflects the divers needs of FM.

• A nationwide environmental scan is underway, engaging 20 to 30 FM organizations selected for their diversity in geography and populations served, as well as AI engagement, gathering insights across all framework domains from leadership and workforce to infrastructure, evaluation, and advocacy.

• Insights will inform updates to the framework, development of an AI playbook, and recognition criteria for FM AI CoEs. These materials will be shared for additional feedback at meetings hosted by the family of FM organizations throughout 2026.

• In 2027, a national FM AI CoE Summit will introduce the first cohort of CoEs and launch a powerful FM AI collaborative to support shared learning, mentorship, and sustained engagement in service to all of FM.

Looking ahead, this effort is not just about building a recognition program. It’s about extending the strong foundations in AI that FM has already built. From early convenings and capacity-building grants to national curriculum development, FM has steadily positioned itself as a leader in responsible innovation. AI is no longer a question of if, but how, who, and for whom it will be used. In FM, our answers must reflect the realities of the communities we serve and the values we’ve long upheld. This ambitious initiative helps us do just that: supporting excellence, fostering innovation, and ensuring that FM is not only present, but leading in the AI era. Together, let us boldly declare as a discipline: family medicine will be the indispensable medical specialty for health care AI.