CLINIC: Evaluating Multilingual Trustworthiness in Language Models for Healthcare

The paper introduces CLINIC, a comprehensive multilingual benchmark designed to evaluate the trustworthiness of language models in healthcare across five dimensions: truthfulness, fairness, safety, privacy, and robustness. Covering 15 languages and 18 tasks, it addresses the lack of reliable evaluation for LMs in low- and mid-resource languages. The study finds that current LMs struggle with factual accuracy, fairness, privacy, and adversarial robustness, highlighting the need for improved global deployment in multilingual healthcare settings.