Minnesota Academy of Family Physicians members are increasingly asking for education on artificial intelligence (AI) in family medicine. The MAFP is committed to providing guidance and support.
MAFP Resolution: Artificial Intelligence in Family Medicine and Primary Care
At the 2025 MAFP Resolution Forum, a resolution affirming the importance of regulation, education and dialogue around AI in family medicine and primary care was adopted.
From the resolution:
- BE IT RESOLVED that the MAFP support regulation and review of AI applications in health care in alignment with the AAFP’s policy Ethical Applications of Artificial Intelligence in Family Medicine.
- BE IT FURTHER RESOLVED that the MAFP work to incorporate AI education and training in MAFP educational and training materials and offerings.
- BE IT FURTHER RESOLVED that the MAFP facilitate discussions among family physicians, family medicine researchers, medical educators, policymakers and technology experts that address issues related to the incorporation of AI tools into the practice of family medicine and primary care in Minnesota.
What Is AI?
Artificial intelligence is an umbrella term for computer-based tools designed by humans to find patterns in large data sets and automate tasks traditionally performed manually.
Types of AI include:
- Machine learning: Finds patterns in smaller or simpler data sets.
- Deep learning: Identifies more complex patterns in large data sets.
- Large language models (LLMs): Generative AI tools such as ChatGPT, Llama, Gemini and Copilot, which predict the “next best word” based on existing data.
Public-facing LLMs should not be used for protected health information or confidential business information, as their data may be used for future model training.
Expert Insights
At the 2025 MAFP Spring Refresher continuing medical education conference, Jason Greenwood, MD, MS, Director of AI and CI, Family Medicine, Mayo Clinic; Anne Scherman, MD, family physician, Allina Health; and Virginia (Ginger) Kakacek, MD, Chief Medical Officer, HealthPartners, led the session AI in Family Medicine: Enhancing Care, Efficiency and Connection, sharing their experiences and challenges with integrating AI into clinical practice.
In addition, the 2025 spring issue of the Minnesota Family Physician magazine, MAFP interviewed David Rushlow, MD, MBOE, Chair of Family Medicine for Mayo Clinic Rochester on his work leveraging AI to improve patient care, streamline clinical workflows and reduce administrative burden. Read the full article.
Key insights can be read below.
How AI Is Used in Clinical Settings
Family physicians are already integrating AI into family medicine in several ways:
- Ambient documentation: Reduces time spent on charting by up to 80% and decreases administrative burden. Some systems integrate directly with electronic health records (EHRs).
- Generative replies: Draft responses to patient messages that can be adjusted for tone, empathy and readability.
- Summarization: Creates concise summaries of visits, hospitalizations or full charts, often improving preparation time and physician confidence before visits.
- Data extraction: Helps identify key details buried in charts and can be linked to clinical decision support or research.
- Predictive models: Forecast outcomes, disease risk or resource needs, though physician input is essential to ensure appropriate development and use.
- Data analysis: Lowers the learning curve for understanding practice data and may reveal new insights.
Only 38% of physicians have implemented the AI scribe services according to The American Academy of Family Physicians (AAFP), even though almost all surveyed expressed interest. The Academy gathered information on eight available AI scribe services, including cost and EHR integration, in partnership with Rock Health, which surveyed family doctors and clinicians to explore AI’s potential and the challenges faced in primary care. More than 65% of respondents reported having limited influence over purchasing decisions regarding AI tools.
Risks and Ethical Concerns with AI
AI is not without risks. “Hallucinations,” or fabricated text, can occur in transcriptions and summaries, raising concerns for clinical accuracy. Researchers have found hallucination rates as high as 80% in some applications. Due to this inaccuracy, editing and source checking must be done.
Bias is another challenge. AI systems are trained on existing data, which can reflect or even amplify biases in race, gender, age or other factors. To minimize this, AI models and scribe services used by clinics must be regularly evaluated, monitored and updated. Family physicians agree that AI needs ethical standards and governance. The AAFP has published ethical guidelines for the use of AI in family medicine.
Evaluating AI in Practice
Recognizing both the potential and the risks of AI, the presenters shared practical strategies for safely integrating these tools into clinical practice.
- Physicians are urged to ask critical questions of vendors, distinguish between true AI and simpler algorithms, and weigh the benefits and risks of adoption.
- Ambient documentation reduced mental burden and improved accuracy, though it did not shorten charting time. Presenters cautioned against AI-generated patient replies that may conflict with physician judgment.
- Insurers are already using AI for prior authorizations, member services and fraud detection, but presenters emphasized the need for human oversight and responsible use.
Across perspectives, presenters agreed: AI will not replace family physicians. When used responsibly, it can reduce burden, enhance patient care and strengthen systems. However, it requires careful evaluation, transparency and ongoing physician advocacy.
You can watch the session on demand for more in-depth information on the topic.
Going Forward
As AI continues to advance, family physicians have a unique opportunity to help shape how these tools are implemented in clinical practice. By staying engaged, asking critical questions and advocating for ethical, patient-centered use, physicians can ensure that AI supports the work of family medicine.
The MAFP will continue to provide resources and forums for discussion, ensuring members feel prepared to navigate this continually evolving landscape.
