AI and the future primary care workforce

ACT Center team members and partners (L-R): Katie Coleman, Rebeckah Muratore, Jess Mogk, and Diane Rittenhouse
With ACT Center and Mathematica, the California Health Care Foundation explores AI technologies in primary care
By Jess Mogk, MPH, a collaborative scientist with the Center for Accelerating Care Transformation (ACT Center) at Kaiser Permanente Washington Health Research Institute
Since early 2024, I've had the pleasure of partnering with Mathematica, former ACT Center Director Katie Coleman, MSPH, and the Healthforce Center at the University of California, San Francisco, on an important project funded by the California Health Care Foundation (CHCF). The project aims to better understand the functions and capacities of primary care teams in California and to inform state policies and investments to strengthen the team-based primary care workforce.
One key role the ACT Center played on this project was contributing to evidence spotlights on several topics of interest to CHCF. There’s been a lot of buzz about artificial intelligence (AI) recently, and we wanted to understand how AI technologies might influence the primary care workforce.
What does AI mean for the future of the primary care workforce?
To explore the implications of AI for the future primary care workforce, our team reviewed relevant literature, attended webinars, interviewed AI thought leaders in primary care, and published our findings in this issue brief available on the CHCF website. Here are a few of the most important insights we gained from this work:
- AI has the potential to reduce administrative burden. At present, many clinicians are responsible for the care of an unreasonably large group of patients, leading to poor work-life balance, exhaustion, and burnout. Clinicians are over-burdened with administrative work, including documentation. Many expect the use of AI will offload a majority of this administrative work and allow primary care teams to focus on building the trusted, healing relationships with patients and families that are foundational to high-quality primary care.
- AI implementation will likely lead to task-shifting on primary care teams, rather than widespread job displacement. Many discussions about AI adoption and the primary care workforce include questions about possible job displacement. In the near term, primary care positions that focus solely on low-complexity tasks (for example, scribing) are most likely to be replaced by AI. Other primary care team members are more likely to see their roles change as AI takes on some tasks. In the context of current workforce shortages, AI should be leveraged to support primary care infrastructure and shore up understaffed teams rather than to replace staff members.
- Generative AI tools have the potential to improve care but must be implemented thoughtfully. One example of how AI is expected to improve care is by being better than busy clinicians at identifying rare conditions. However, improvements in care quality are not a guarantee in every situation, and guardrails should be put in place to provide oversight and prevent harm. In addition, there may be some tasks AI should not manage, and AI cannot be expected to replace the trusted human relationships that are essential to high-quality primary care.
- To leverage AI, the primary care workforce of the future will need technical skills and support. Primary care teams should receive training on how to work with AI, including instruction on its capabilities and limitations, and how to identify AI failures. Teams will need adequate support to address technical problems when they arise.
In the context of primary care, AI should be considered a tool and not be expected to fix the multifaceted workforce issues that have developed over decades due to misaligned incentives in the U.S. health care system. While it’s too soon to predict exactly how AI will change the primary care workforce, it’s almost certain that care teams will need to learn how to work with AI. Primary care teams will need adequate support to adopt these new technologies in a way that doesn’t increase burden. Leaders implementing AI should think carefully about which tasks will be managed by AI versus staff and how AI performance will be monitored. As AI tools evolve and become more widespread and digital natives enter the primary care team workforce, the nature of primary care work could look very different in the future.
We look forward to publishing and sharing more findings from this project soon, especially our policy recommendations to optimize the primary care team workforce in California. Please stay tuned and visit the CHCF website to access more resources on bolstering the health care workforce.