Comparison of AI-generated renal diets by different large language models: a guideline-based evaluation with expert input

Key Points

• AI-generated renal diets show notable differences based on the language model used, indicating model biases.
• The assessment included evaluations from several language models, comparing dietary recommendations for renal patients.
• Guideline-based evaluation revealed that only specific models aligned well with established dietary guidelines for renal health.
• Implications of this work may influence how AI models are trained to generate clinically relevant dietary advice.