Large language models (LLMs) offer substantial promise for improving health care; however, some risks warrant evaluation and discussion. This study assessed the effectiveness of safeguards in foundational LLMs against malicious instruction into health disinformation…