Healthcare AI—Current Applications and What’s Next: An Expert Panel Discussion

October 2025

Author

Ronald Poon Affat, FSA, FIA, MAAA, CFA, HIBA

Executive Summary

The basis for this expert panel report came from the presentation at the NAIC’s 2025 Spring National Meeting, where newly released survey results noted the accelerating adoption of artificial intelligence in insurance and health care. The NAIC’s survey of 93 health insurers revealed that 92% are already using, planning to use, or actively exploring AI. This aligns with broader NAIC findings across four major lines of business—health, auto, home, and life—with health leading at 92%, followed by auto at 88%, home at 70%, and life at 58%.

The NAIC survey makes clear that health insurers have been early leaders in adopting AI, applying it to areas such as claims, fraud detection, and customer service. The same transformative potential is now taking shape on the provider side, where hospitals and clinics—vital pillars of the healthcare ecosystem—are beginning to use AI not only for administrative efficiency but also in core aspects of patient care. Although adoption remains at an early stage, the trajectory points toward convergence between payer and provider applications, with shared opportunities and risks requiring coordinated approaches. Against this backdrop, the panel focused on how hospitals are integrating AI into clinical and operational workflows.

The panel brought together a uniquely diverse set of professionals spanning actuarial science, healthcare data science, enterprise analytics, and clinical AI governance. Their combined backgrounds encompassed research on emerging risks, deployment of predictive modeling in healthcare payer–provider settings, the creation of enterprise data strategies with rigorous governance, hands-on implementation of clinical AI oversight, and operationalization of generative AI and natural language processing tools. This breadth of expertise allowed for a multi-dimensional examination of both the opportunities and risks in integrating AI into hospital workflows.

Discussions underscored AI’s measurable benefits in improving equity and efficiency. Examples included AI-assisted diagnostic tools that reduced delays in radiology workflows, models that accelerated triage accuracy, and predictive systems identifying high-risk patients earlier—often benefiting underserved communities. The panel also explored the challenges of potential bias, noting how seemingly neutral data points, such as geographic markers, can become proxies for protected attributes, risking inequitable outcomes if left unmitigated. Strategies to address these risks included embedding fairness audits into development cycles, expanding and diversifying training datasets, and establishing multi-disciplinary review processes that integrate actuarial, clinical, and technical perspectives.

A recurring theme was that the success of AI adoption depends as much on organizational readiness as on technical capability. Hospitals that achieved sustained results invested in staff education, transparent governance frameworks, and clear escalation protocols for situations where human and AI recommendations diverged. These governance structures, coupled with continuous monitoring, ensured AI tools were trusted and used consistently in practice.

Material

Healthcare AI: An Expert Panel Discussion

Acknowledgements

The researchers’ deepest gratitude goes to those without whose efforts this project could not have come to fruition: the expert panel participants and others for their diligent work overseeing questionnaire development, analyzing, and discussing respondent answers, and reviewing and editing this report for accuracy and relevance.

Dale Hall, FSA, MAAA, CFA, CERA, Managing Director of Research at the Society of Actuaries; specializes in insurance research, actuarial science, and public policy.

Joe Dorocak ASA, MAAA Director, Enterprise BI at Medical Mutual

Joe Long ASA, MAAA, Consulting Actuary & Data Scientist, Milliman; specializes in healthcare AI, predictive modeling, and data infrastructure

Colleen Houlahan, Data Scientist, Cleveland Clinic; part of the Artificial Intelligence Operations team implementing clinical AI governance.

Jacob Raciniewski, VP of Enterprise Data & Analytics, Medical Mutual; oversees enterprise analytics strategy, including AI adoption for claims processing and member engagement.

Ronald Poon Affat, FSA, FIA, MAAA, CFA, Independent Board Director and Cross Continental Actuary; expert in AI governance, reinsurance and emerging market insurance.

At the Society of Actuaries Research Institute:

Korrel Crawford, Senior Research Administrator

Questions or Comments?

Give us your feedback! Take a short survey on this report. Take Survey

If you have comments or questions, please send an email to Research@soa.org