Publications | TIERRA

A person in a yellow sweater selects a book from a wooden shelf in a library.

Exploring Ethical AI in Healthcare: Insights, Research, and Policy Perspectives

Highlights

Drawing on a community deliberation with Arab/MENA participants in Michigan, this article explores hopes and concerns about healthcare AI, including transparency, cultural representation, human connection, and the role of providers. It outlines practical implications for how health systems can build trust and implement AI in ways that support clinicians while reflecting the needs and values of underrepresented communities.

  • R.Hamasha, D.Saleem, K. A.Ryan, et al., “Engaging Arab/MENA Communities in Learning Health Systems: Insights and Guidance for Future Research on AI and Health Equity,” Learning Health Systems10, no. S1 (2026): e70091, https://doi-org.proxy.lib.umich.edu/10.1002/lrh2.70091.

Drawing on community deliberations, this article explores public hopes and concerns about health AI including, transparency, privacy, equity, safety, and accountability, and outlines practical implications for how health systems can design and implement trustworthy AI that aligns with community values.

  • Ryan KA, Sielaff ML, Saleem D, Richardson J, Tan S, Hamasha R, Nong P, Kardia SLR, Romanov V, Hammad A, Platt J. Community perspectives on health AI: hopes, concerns and implications for health systems and trustworthy AI. AI Ethics. 2026;6(2):176. doi: 10.1007/s43681-026-00987-7. Epub 2026 Feb 26. PMID: 41768100; PMCID: PMC12945899.

A Michigan-based deliberative study found strong public support for patient-informed artificial intelligence (AI) labeling in health care, emphasizing transparency, privacy, equity, and safety to build trust.

  • Sielaff ML, Platt J, Tan S, Ryan KA, Nong P, Kardia SLR. Building trust: public priorities for health care AI labeling. Am J Manag Care. 2026 Jan 1;32(1):e18-e24. doi: 10.37765/ajmc.2026.89875. PMID: 41592186.

The study finds that while large language models, like ChatGPT, are useful for medical free-text classification, alternative models can offer better accuracy and lower costs. It highlights ongoing challenges such as computational expenses and reliability, suggesting that a mix of local and commercial models may provide the best value for healthcare systems.

A national survey found that most US adults have low trust in their health care system to use AI responsibly and protect them from AI-related harms, highlighting the need for greater patient-centered approaches as AI use in health care grows.

Michigan Medicine researchers found that patients overwhelmingly prefer direct communication about how their health data and biospecimens are used in research, highlighting the importance of transparency and respect for patient preferences.  

  • Spector-Bagdady K, Ryan KA, Chen L, et al. Lessons for a learning health system: Effectively communicating to patients about research with their health information and biospecimens. Learn Health Sys. 2025; 9(1):e10450. doi:10.1002/lrh2.10450

Nong, P., Adler-Milstein, J., & Platt, J. (2024). How patients distinguish between clinical and administrative predictive models in health care. The American journal of managed care, 30(1), 31–37. https://doi.org/10.37765/ajmc.2024.89484

Nong, P., Hamasha, R., & Platt, J. (2024). Equity and AI Governance at Academic Medical Centers. 30, SP468–SP472.

Cross FL, Hunt R, Buyuktur AG, et al. Factors That Impact Effective Public Health Communication With Michigan’s Latinx Population in the Context of COVID-19. Health Education & Behavior. 2024;52(1):82-91. doi:10.1177/10901981241278962

Nong, P., Hamasha, R., Singh, K., Adler-Milstein, J., & Platt, J. (2024). How Academic Medical Centers Govern AI Prediction Tools in the Context of Uncertainty and Evolving Regulation. NEJM AI, 1(3), AIp2300048. https://doi.org/10.1056/AIp2300048

Platt J, Nong P, Carmona G, Kardia S. Public Attitudes Toward Notification of Use of Artificial Intelligence in Health Care. JAMA Netw Open. 2024;7(12):e2450102. doi:10.1001/jamanetworkopen.2024.50102

Jodyn Platt, Paige Nong, Renée Smiddy, Reema Hamasha, Gloria Carmona Clavijo, Joshua Richardson, Sharon L R Kardia, Public comfort with the use of ChatGPT and expectations for healthcare, Journal of the American Medical Informatics Association, Volume 31, Issue 9, September 2024, Pages 1976–1982, https://doi.org/10.1093/jamia/ocae164

Paige Nong, Julia Adler-Milstein, Sharon Kardia, Jodyn Platt, Public perspectives on the use of different data types for prediction in healthcare, Journal of the American Medical Informatics Association, Volume 31, Issue 4, April 2024, Pages 893–900, https://doi.org/10.1093/jamia/ocae009