Introduction
Generative AI (genAI) is rapidly entering healthcare, but public comfort may vary in ways that could undermine equitable adoption.
Methods
Using two independent cross-sectional survey waves of US adults conducted in May 2023 (N = 3130) and November 2025 (N = 3088) on the Outbreaks Near Me participatory surveillance platform, we assessed changes in awareness of genAI and four comfort and perception measures.
Results
The percentage of respondents reporting they had heard “a lot” about genAI rose from 26.5% (95% confidence interval [CI]: 24.5, 28.5) to 42.9% (95% CI: 40.6, 45.3), yet overall comfort with AI-led primary care and AI-led therapy changed minimally (-1.0 and -0.7% points, respectively). Subgroup trends diverged sharply by political identity, age, and race. Notably, Republicans trended more favorably from 2023 to 2025 while Democrats declined (interaction odds ratios across the four measures ranging from 1.69 [95% CI: 1.19, 2.39] to 2.11 [95% CI: 1.54, 2.90]).
Conclusion
Emerging divides suggest receptivity to genAI in healthcare may become increasingly shaped by political identity rather than clinical evidence or personal experience, complicating equitable deployment.