Food waste in grocery retail generates 8-10% of global greenhouse gas emissions, and AI is attacking this problem at three points: pre-delivery demand forecasting, shelf-level dynamic pricing, and delivery route optimization. ML models that factor in weather, promotions, and local events can reduce shrink by 20-25%. Smart shelves with expiry-triggered discounts (as demonstrated at a London independent supermarket) save hundreds of products weekly and cut CO2 emissions. For e-commerce, AI consolidates delivery routes and right-sizes packaging. The piece also addresses real barriers: manager overrides of AI recommendations, poor inventory data quality, and AI's own energy footprint. The core argument is that waste reduction is primarily an information problem, and the business case (not sustainability pledges) is what drives adoption.
