Unhealthy diet is a key driver of the growing global burden of metabolic diseases, such as type 2 diabetes and obesity. Dietary change is therefore a cornerstone of diabetes prevention and a central component of its management, but its potential remains underutilized. Making the right food choices in everyday life is inherently challenging: many people lack basic knowledge about food, are constrained by entrenched habits, face cultural or financial barriers, and are overwhelmed by the abundance of food choices and misleading health claims. 

Although dietary counseling by qualified professionals is effective, it is limited by consultation time, insight into real-world eating behaviors, and a lack of tools to provide truly personalized, iterative guidance tailored to individual adherence patterns and health goals. Digital technologies, wearable devices, and generative AI hold substantial potential to bridge these gaps by providing scalable, data-driven, and individualized nutrition support. Integrating continuous glucose monitoring, digital diet tracking, and individual context (e.g., BMI, food preferences, budget) enables generative AI to provide context-aware counselling and personalized meal suggestions.

However, this field remains nascent, and AI-enabled personalized nutrition must be rigorously developed and validated in close collaboration with nutritional experts to ensure recommendations are safe, nutritionally adequate, and clinically meaningful.

The AI-Nutritionist project evaluates the efficacy of automated, AI-driven dietary counselling compared to standard-of-care dietary counselling in patients with or at risk of type 2 diabetes.

The goal is for every patient to have access to validated, safe, and nutritionally appropriate dietary advice anytime, anywhere.