Researchers from the University of California, in collaboration with Microsoft, are developing an advanced AI tool designed to help communities anticipate crop failures and mitigate the impact of malnutrition.

The initiative comes in response to the severe drought that affected Kenya in 2023, leaving 4.4 million people facing acute food insecurity and 1.1 million women and children suffering from acute malnutrition, according to the UN’s World Food Programme.

To enhance early intervention, researchers have analyzed weather patterns, NASA satellite imagery, and clinical health records from the past decade. These data sources have been integrated into computational models capable of predicting food insecurity trends up to six months in advance.

AMREF, one of the organizations supporting the project, has been trialing the AI tool in Kenya since February 2024 year. While exact figures are yet to be released, early results indicate high predictive accuracy over three- to six-month periods.

Samuel Mburu, an AMREF expert involved in the project, emphasized the tool’s potential to allocate resources more efficiently. “By analyzing historical data at the sub-county level, we can ensure that vulnerable children under five receive the necessary nutrition at health facilities before a crisis escalates,” he explained.

Initially focused on six drought-prone counties in northern and eastern Kenya, the model has now expanded to cover the entire country. AMREF highlights that relying solely on health records is insufficient, as many affected individuals lack access to clinical services. In the future, the organization aims to incorporate data from community sources to improve accuracy.

Beyond prediction, the AI tool is expected to enhance resource mobilization by providing concrete data to attract funding. “Previously, data was used for historical reporting, but now we can use it to forecast needs and secure resources both locally and internationally,” Mburu noted.

While the project benefits from Microsoft’s sponsorship, cost remains a challenge, particularly in cloud-based environments where AI processing demands substantial resources.

Local county officials are also set to benefit from the technology, using its forecasts to plan ahead for food insecurity. “Instead of reacting to malnutrition cases as they arise, this tool allows us to predict trends and mobilize food supplies in advance, ensuring timely intervention for at-risk communities,” Mburu added.

This AI-driven approach marks a significant step toward combating food insecurity in Kenya, offering a proactive solution to protect vulnerable populations from the devastating effects of climate-related food shortages.