US poultry industry develops real-time avian influenza risk platform using machine learning

In response to recurrent outbreaks, the US poultry sector is investing in an innovative open-source platform that uses data and AI to forecast avian influenza risks, aiming to shift from reactionary measures to proactive prevention.

Avian influenza remains one of the poultry sector’s most persistent operational risks, and the 2022 outbreak showed how quickly it can unsettle supply chains. According to USDA Economic Research Service data, egg inventories fell 29% by December 2022 from the start of the year, more than 43 million laying hens were lost and wholesale egg prices surged to record levels. The same outbreak had a more limited effect on broiler production, but it still removed 2.3 million birds, while turkey producers saw around 5.4 million birds depopulated.

Against that backdrop, the U.S. Poultry & Egg Association has funded University of Georgia research to build an open-source platform that would forecast avian influenza risk across the US in near real time. The project aims to move the industry from a largely reactive posture towards earlier intervention, using data on bird migration, weather, wetlands, poultry density and historical outbreaks.

The system is being designed to combine statistical modelling with machine learning so producers and veterinarians can identify hotspots, strengthen biosecurity and direct surveillance more efficiently. For an industry exposed to disease shocks and trade disruption, the value lies not in prediction alone, but in enabling faster, better-informed decisions before outbreaks spread.