J Med Internet Res. 2026 Apr 29;28:e86769. doi: 10.2196/86769.
ABSTRACT
This study explores the role of open-source large language models (LLMs) in promoting artificial intelligence (AI) health equity from the perspective of the health service triangle model. First, it defines AI health, categorizes AI-supported decision-making patterns, and assesses the status quo of AI health inequalities. Second, by comparing open-source and closed-source LLMs in terms of patient privacy, data security, accessibility, and use, it demonstrates the distinct advantages of open-source LLMs for AI-enabled health services. Finally, based on the health service triangle model, this study demonstrates how open-source LLMs drive the democratization of AI-enabled health services-particularly benefiting low-resource regions-by expanding service types, improving accessibility, enhancing quality, and reducing costs. This study concludes that, while open-source LLMs must address challenges such as hallucination risks and ethical responsibilities, they ultimately enable AI health equity through technological sharing.
PMID:42054668 | DOI:10.2196/86769
