What Does the AI Doctor Value? Auditing Pluralism in the Clinical Ethics of Language Models
A new study has developed a framework to audit the ethical values embedded in large language models used for medical advice. The research found that while frontier models exhibit a range of ethical priorities similar to human physicians, their decisions are often deterministic and may underweight patient autonomy. This could lead to a "deployment monoculture" of ethical perspectives if not addressed, potentially replacing the pluralism inherent in clinical practice. AI
IMPACT AI medical advice tools risk imposing a narrow ethical framework, potentially undermining patient autonomy and clinical pluralism.