Challenges of Auditing: Variability in Outputs of Large Language Models for Health
Not on the map. Why?
Abstract (arXiv)
People increasingly use frontier AI models for health advice, but via different access modes (e.g., ChatGPT, ChatGPT Health, APIs) with varying settings. Here, we find systematic differences across access modes. Because evaluations typically rely on APIs while consumers interact through chatbot interfaces, these discrepancies limit evaluation validity. Our findings underscore an urgent need for model providers to enable faithful replication of consumer experiences and settings for rigorous audits.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 9 of its 22 references that are in the map. First posted on arXiv 15 Sep 2026.
Builds on
9 of 22 references.
References · 9
- 2025HealthBench: Evaluating Large Language Models Towards Improved Human Healthcited by 429
- 2023ChatDoctor: A Medical Chat Model Fine-Tuned on a Large Language Model Meta-AI (LLaMA) Using Medical Domain Knowledgecited by 325
- 2024Black-Box Access is Insufficient for Rigorous AI Auditscited by 126
- 2023Better to Ask in English: Cross-Lingual Evaluation of Large Language Models for Healthcare Queriescited by 82
- 2025Large language models provide unsafe answers to patient-posed medical questionscited by 18
- 2024An Empirical Evaluation of Large Language Models on Consumer Health Questionscited by 6
- 2026LLM Spirals of Delusion: A Benchmarking Audit Study of AI Chatbot Interfacescited by 6
- 2024Evaluating Search Engines and Large Language Models for Answering Health Questionscited by 5
- 2026What Current AI Benchmarks Leave Unmeasured: Modality, Search, Citations, and Implications (for Safety Evaluations)cited by 1
Built on by
Not cited yet.