Return to Article Details When fluent advice becomes unsafe: evaluating accuracy, uncertainty, and communicative risk in Indonesian AI-generated health responses Download Download PDF