With Patients in Mind …

Leading health systems are implementing artificial Intelligence (AI) to increase operational efficiencies, lower costs and improve outcomes. In fact, generative AI (GenAI) is already changing how healthcare organizations analyze data, manage care and streamline administrative tasks.

As healthcare systems implement more AI technology to improve operations, outcomes and patient experience, actuaries are serving as strategic partners within healthcare systems. By applying their expertise in model governance and scenario planning while collaborating across payer-provider boundaries, they can help build trust in AI-driven healthcare.

REAL-LIFE AI HEALTHCARE APPLICATIONS

Whether it’s accurately answering the questions of first-time mothers in real time, helping caregivers navigate the complexities of healthcare or facilitating patients’ follow-up appointments, AI has the potential to improve patient care. The SOA Research Institute report, Provider Use of AI in Healthcare, illustrates a number of real-life applications of AI use cases in prominent healthcare systems.

Patient risk stratification is a classic example of the role predictive analytics can play in healthcare. Because electronic medical records (EMRs) store patients’ medical history, they are comprehensive datasets that can enable predictive models to help providers assess a patient’s risk profile. For example, a patient with elevated heart rate variability, low blood pressure, abnormal lab results and signs of frailty could be identified as high risk, prompting an early conversation about their care needs. Timely intervention and discharge to the appropriate care setting can improve outcomes while creating a more empathetic and supportive patient experience.

Additionally, hospital systems can forecast key metrics to more effectively plan resource allocation and operational efficiency. Due to the complexity of healthcare systems, this kind of forecasting has been difficult historically, but advancements in methodologies provide the potential to improve the precision of predictions and enhance hospital management.

For example, demand-first forecasting enables healthcare systems to identify patterns of emergency department visits and use this information to ensure sufficient staffing during busy times. In addition, supply-first forecasting can help healthcare managers schedule office visits and operating room utilization in concert with availability of staff and resources.

Prescriptive analytics is another AI-enhanced tool helping healthcare systems improve decision-making and, ultimately, outcomes. By incorporating potential interventions into predictive models as additional covariates, providers can gain actionable insights. For example, a model predicting a 30-day hospital readmission risk might rank follow-up texts, special visits and prescription changes by their potential to reduce that risk compared with no intervention. These estimates could help providers tailor care and allocate resources to improve patient outcomes.

Telemedicine and EMR interactions have become two of the more prominent uses of GenAI in healthcare. AI can assist in reviewing patient messages and charts to draft relevant replies, helping providers respond faster and capture key information.

NEEDS: GUARDRAILS, DATA AND TALENT

At a November 2025 roundtable discussion sponsored by the SOA Research Institute, actuaries from consulting firms and health insurers stressed that clear AI risk management and governance help organizations innovate and scale responsibly. That means defining roles, responsibilities and escalation paths while aligning with existing risk management and compliance frameworks.

Furthermore, Provider Use of AI in Healthcare finds that, as AI becomes part of everyday care, healthcare organizations need to address potential bias in algorithms, protect patient privacy, ensure accountability and transparency, and inform patients about AI’s benefits and risks. Healthcare organizations will need to navigate a complex landscape in which they are required to stay in tune with emerging regulations, provide transparency in their work with AI and earn patients’ trust while embracing innovation.

Healthcare systems also need high-quality patient data to realize AI’s potential. Cleveland Clinic’s use of quantum computing for molecular modeling, drug discovery and personalized medicine shows what advanced technology can offer when it is heavily trained on historical data.

Finally, healthcare providers will aspire to find the talent and skills needed to build and deploy AI at scale. Many are building in-house expertise and resources to develop these complex systems.

POSSIBLE SOLUTIONS: INNOVATION, COLLABORATION AND ACTUARIES

Some of the solutions to issues are already available. The November 2025 roundtable participants noted that actuaries’ experience with model governance, documentation and regulators makes them well-suited to contributing to governance development.

To build patient and provider trust in AI and provide transparency to regulators, healthcare systems may form governance teams that bring together actuaries, legal, compliance, data science and clinical staff, according to Provider Use of AI in Healthcare. These teams can guide AI use and set standards for validating and monitoring applications.

To overcome the challenge of gaining access to patient data, health systems can leverage open-sourced information. For example, researchers and developers can mine vast amounts of medical literature to refine their algorithms. Data twinning and synthetic data generation can create large datasets without compromising sensitive patient information, reducing patient privacy concerns.

Additionally, hospitals and other providers can partner with technology companies to gain the specialized skills and talent needed to build and deploy AI systems at scale.

Realizing AI’s potential in healthcare will require sound governance, access to data and collaboration among clinicians, technology experts and payers. Actuaries are already connected with these efforts by evaluating risk, testing assumptions and assessing whether AI applications deliver measurable value with the opportunity to do even more in the future. By bringing their expertise to multidisciplinary teams, they can help healthcare organizations implement scalable solutions that improve operations, support better care and earn patients’ trust.

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries or the respective authors’ employers.

Copyright © 2026 by the Society of Actuaries, Chicago, Illinois.