Sustainable Bites Leveraging AI to Decode Food Emissions and Nutrition for Conscious Consumption

The environmental footprint of global food produc- tion accounts for approximately 26 percent of total greenhouse gas (GHG) emissions, making dietary choices a critical factor in climate change mitigation. However, assessing the environmental and nutritional impact of processed food products remains a challenge due to proprietary recipe information and complex supply chains. This study presents an AI-driven framework that utilizes open-source large language models (LLMs) to approximate ingredient compositions of various food products. By integrating these approximations with emissions databases and an updated NutriScore system, we develop a comprehensive dataset that provides both GHG impact and nutritional value for each product. The data is stored in a queryable vector database, allowing consumers to identify lower-impact, healthier alterna- tives in real time. This research demonstrates the potential of AI to bridge knowledge gaps in food sustainability and empower consumers with data-driven insights for making environmentally responsible and health-conscious dietary decisions.