AI models offer healthier, cheaper meals but raise misinformation concerns

Researchers at the University of California, Davis created a generative AI system that can suggest one to three ingredient substitutions to make meals align more closely with USDA nutrition targets. In simulations the AI‑generated meals were 47% nearer to dietary guidelines, improved nutritional quality by about 10% and reduced estimated costs by 22%‑34%.

Nutrition experts caution that AI‑generated dietary advice can be unreliable. Analyses suggest as many as 50% of such recommendations may be inaccurate or misleading because large language models lack clinical judgment and cannot account for individual medical histories. Misguided AI advice could lead to nutrient deficiencies, mismanagement of chronic conditions and higher health‑care costs, prompting professionals to advise users to verify any AI‑derived diet plan with qualified dietitians or physicians.