Aggregate benchmark scores obscure patient safety implications of errors across frontier language models
Preprint
Abstract
Frontier language models are widely used for health-related queries, yet aggregate benchmark scores do not capture safety implications of errors. We applied the recent Nature Medicine triage benchmark across nine frontier models, comparing directional error profiles, contextual bias, and crisis calibration. In-range accuracy ranged from 75.0% to 87.7%, obscuring clinically meaningful error differences. Looking at directionality of errors, under-triage ranged from 0.0% (GPT-5.2) to 12.3% (GPT-5-mini), over-triage varied independently (9.4–36.9%), and under-triage was uncorrelated with aggregate accuracy. When family members minimized symptoms, all models tested shifted toward lower acuity in ambiguous cases (OR range 2.9–14.9), the only contextual effect observed consistently, and access barriers increased under-triage risk in six. Suicide crisis resource mention rates were low and variable across all models. This cross-model heterogeneity and non-monotonic performance across model generations show that aggregate accuracy alone cannot characterize, rank, or predict the clinical safety of deployed language models.Keywords
This publication has 11 references indexed in Scilit:
- ChatGPT Health performance in a structured test of triage recommendationsNature Medicine, 2026
- Online Health Information–Seeking in the Era of Large Language Models: Cross-Sectional Web-Based Survey StudyJournal of Medical Internet Research, 2025
- Laypeople’s Use of and Attitudes Toward Large Language Models and Search Engines for Health Queries: Survey StudyJournal of Medical Internet Research, 2025
- An evaluation framework for clinical use of large language models in patient interaction tasksNature Medicine, 2025
- The Triage and Diagnostic Accuracy of Frontier Large Language Models: Updated Comparison to Physician PerformanceJournal of Medical Internet Research, 2024
- Characterizing the Adoption and Experiences of Users of Artificial Intelligence–Generated Health Information in the United States: Cross-Sectional Questionnaire StudyJournal of Medical Internet Research, 2024
- The diagnostic and triage accuracy of the GPT-3 artificial intelligence model: an observational studyThe Lancet Digital Health, 2024
- Evaluation and mitigation of the limitations of large language models in clinical decision-makingNature Medicine, 2024
- The shaky foundations of large language models and foundation models for electronic health recordsnpj Digital Medicine, 2023
- Probable Inference, the Law of Succession, and Statistical InferenceJournal of the American Statistical Association, 1927