Nutrition plays a pivotal role in the prevention and management of non-communicable diseases (NCDs), including obesity, type 2 diabetes, cardiovascular diseases, and certain types of cancer, which represent a major burden on healthcare systems and significantly impact quality of life. In this context, front-of-pack (FOP) labelling has emerged as a critical policy tool to encourage healthier dietary choices by providing understandable nutritional information at the point of purchase.
However, despite existing FOP schemes (e.g. Nutri-Score, Nutri-Form Battery), challenges persist. The heterogeneity of labelling systems across European Member States, limited standardisation of nutrient profiling criteria, and widespread low health literacy often limit the effectiveness of these tools in driving meaningful behavioural change. Complex ingredient lists, non-uniform label designs, and culturally specific dietary patterns contribute to consumer confusion and reduce the usability of current labels.
In this evolving policy landscape, Artificial Intelligence (AI) offers a transformative opportunity to strengthen FOP labelling systems and improve public health outcomes. In fact, through the integration of machine learning, natural language processing (NLP), computer vision, and real-time analytics, AI can enhance the transparency, personalisation, and impact of nutrition information in several key areas.
- Standardisation and Decoding of Label Data. AI algorithms can decode and standardise unstructured label content across diverse packaging formats, languages, and typographies. Using computer vision and OCR (optical character recognition), AI tools can extract detailed nutritional data from product images, enabling automatic classification of food items based on nutrient profiles, levels of processing, and health claims. This can facilitate cross-country comparability and support regulatory harmonisation across the EU single market.
- Nutritional Profiling and Risk Stratification. AI-based models can evaluate the nutritional quality of food products by aligning with established nutritional WHO/EFSA guidelines. Moreover, machine learning can integrate additional dimensions such as portion size, food matrix, and context of consumption, to offer a more nuanced interpretation of a product’s health impact. These evaluations can be scaled to large product databases, enabling ongoing surveillance of market trends and formulation shifts.
- Personalised Dietary Feedback. One of the most promising applications of AI in nutrition labelling lies in personalisation. Using user-generated data (e.g. age, BMI, allergies, disease predisposition, microbiota composition, or wearable sensor outputs), AI tools can deliver tailored feedback and alerts. For instance, individuals with hypertension can receive warnings on high-sodium products, while those at risk of metabolic syndrome may be guided toward low glycemic-index alternatives. Such personalised digital labelling has the potential to shift dietary choices from generic advice to targeted interventions.
- Dynamic and Context-Aware Labelling. AI also supports the development of dynamic labelling systems that adjust in real time. By leveraging cloud-based platforms and user health data, FOP labels could become interactive and adaptive. Integration with mobile applications, smart fridges, or augmented reality (AR) interfaces can provide users with contextualised recommendations based on recent behaviours, nutrient deficits, or even epigenetic risk factors. This aligns with the emerging field of precision nutrition and its potential to enhance metabolic health and disease prevention.
- Applications and Pilot Programs. Several AI-powered platforms are already demonstrating the feasibility and benefits of this approach. Tools and applications offer barcode scanning features that provide immediate product ratings based on health and environmental criteria. These tools often use crowdsourced or manufacturer-submitted data, enriched through AI-driven analysis, to offer consumers transparency and actionable insights. Besides, at the institutional level, pilot studies are currently assessing the effectiveness of AI-enhanced digital labels in school canteens, hospital cafeterias, and supermarket settings. Early findings suggest improved nutritional literacy, higher compliance with dietary recommendations, and measurable shifts in purchasing behaviours, especially when feedback is tailored to individual or group needs.
- Policy Monitoring and Compliance. Beyond consumer-facing tools, AI also offers value for regulatory bodies. Algorithms can be trained to detect false or non-compliant health claims, track the marketing of ultra-processed products, and evaluate adherence to food reformulation targets. This automated surveillance can support enforcement actions, ensure transparency, and accelerate the implementation of the Farm to Fork Strategy’s objectives.
Therefore, by embedding AI within nutrition labelling systems, the EU can move toward a model of precision public health, one that aligns digital innovation with policy coherence and consumer empowerment. AI can help bridge the gap between scientific evidence and consumer behaviour, translating complex data into accessible, actionable, and personalised guidance. In doing so, the Union can better address the burden of NCDs, promote equitable access to healthy diets, and support the transition to a sustainable and resilient food system. As we approach the implementation milestones of the Farm to Fork Strategy, AI should not be viewed as a substitute for regulation but as an enabler of smarter, more inclusive, and more effective policy execution.
References
- European Commission. “A Farm to Fork Strategy for a fair, healthy and environmentally friendly food system”, COM(2020) 381 final. https://food.ec.europa.eu
- Weber et al. Journal of Industrial and Business Economics. https://doi.org/10.1007/s40812-025-00355-2
- Batista MF et al. Front-of-package nutrition labeling as a driver for healthier food choices: Lessons learned and future perspectives. Compr Rev Food Sci Food Saf. 2023;22(1):535-586. doi:10.1111/1541-4337.13085
- Peonides M et al. Food labeling in the European Union: a review of existing approaches. International Journal of Health Governance. 2022;27(4):460-468. https://doi.org/10.1108/IJHG-07-2022-0072
- Kelly, B. The potential effectiveness of front-of-pack nutrition labeling for improving population diets. Rev. Nutr.2024;44:405–440
- AI-FOOD: AI-based mobile app: https://www.eitfood.eu/projects/ai-food-ai-based-mobile-app
Kim D, et al. Innovative AI methods for monitoring front-of-package information: A case study on infant foods. PLoS One. 2024;19(5):e0303083. Published 2024 May 16. doi:10.1371/journal.pone.0303083
