AI-driven tools aim to spot and stop zoonotic disease spillovers before they spark pandemics

Conservation Through Public Health collected samples from livestock around Bwindi Impenetrable National Park and sent them to a lab in Entebbe, where diseases like brucellosis and Rift Valley fever were identified. The results were merged with gorilla-monitoring data and used by the NESTLER initiative to train predictive AI models that act as early-warning systems for humans, livestock and endangered gorillas. Across the globe, scientists such as Edward Holmes are training deep-learning algorithms on public viral sequence databases to spot “risky” viruses, while firms like BlueDot employ AI to scan news, travel and climate data for emerging threats.

HealthMap, now powered by large-language models, and the WHO’s pandemic-intelligence hub are also part of a growing network of AI-enhanced surveillance. Although the technology shows promise, experts warn that bias and the need for human oversight remain challenges to reliable outbreak prediction.