In Somalia, AI is helping deliver food to hungry families

Although the UN’s World Food Programme (WFP) cannot reach all 6 million Somalis facing hunger, data collected by its partners and analysed through AI tools in its hunger-tracking platform HungerMap Live help the agency position resources in specific areas before conditions deteriorate.

Somalia faces climate change-induced droughts followed by flash floods and a brutal conflict that has persisted for 35 years and killed hundreds of thousands.

The Horn of Africa country also imports most of its food, exposing it to global price shocks such as the current conflict in the Middle East.

In June, WFP and the UN’s Food and Agriculture Organization (FAO) added Somalia to their list of “hunger hotspots of highest concern” alongside northeast Nigeria, joining Sudan, South Sudan, Yemen and Palestine.

“It is a very gruesome situation, and I’m very, very concerned about these next few weeks and months,” said Simon Renk, head of vulnerability, analysis and mapping for the WFP in Somalia.

Using AI in Burhakaba

In the Burhakaba District, about 180 kilometres northwest of the capital, Mogadishu, dangerously malnourished children fill local medical centres.

WFP’s humanitarian workers identified the district’s urgent need using the AI tools in HungerMap Live.

Armed with that knowledge, the agency provided food assistance to 48,000 people and nutrition support to 3,000 women and children.

“Instead of waiting for that malnutrition of children to deteriorate and lead to displacement or even death, we knew it prior,” Mr. Renk told UN News.

Old technology, new lifesaving potential

First launched in January 2020 to monitor food security in more than 90 countries, HungerMap Live is a publicly-accessible platform that uses data on food insecurity, climate, economies, agriculture, inflation and more to display areas that are currently food insecure or moving in that direction.

Although that initial version of HungerMap Live used AI to estimate levels of food insecurity in countries lacking sufficient data, its ability to predict future hunger crises was limited and it did not track the nutritional quality of diets in the countries it tracks.

Enter HungerMap Live 2.0. Released by WFP in April, the update overhauled the user interface and introduced new predictive AI features, allowing governments and WFP field offices around the world to mobilise resources and act before a crisis worsens.

“With this new version, it helps us not only to describe where hunger is but also helps us to explain why conditions are deteriorating and where they could worsen next,” Mr. Renk said.

While the data from HungerMap Live cannot replace on-the-ground assessment, it helps the WFP team in Somalia determine which communities to help first and which type of aid is most helpful to a particular community – whether it be cash assistance or specialised assistance to address malnutrition.

Prioritising aid in Somalia

In a country where 6 million people experience crisis levels of food insecurity, the WFP is only able to reach one in 10 people in need.

WFP’s office in Somalia requires an additional $192 million between now and January 2027 to reach everyone who needs food, Mr. Renk said.

With resources stretched thin, he said AI-powered tools help the agency decide which communities to prioritise and help cut the time between early warning and lifesaving action.

Using AI, WFP analyses predictive data on rainfall deficits, flood risk, market prices, conflict and nutrition around the world to forecast where hunger will strike next – allowing governments to take preventive action and WFP and its partners to pre-position crucial assistance.

Time is of the essence

“This platform and different solutions are not meant to make hunger more technical or abstract. What I see is fundamentally to shorten the time between an early warning and a real decision in order to protect a family and individual before it really becomes irreversible,” said Simon Renk, head of vulnerability, analysis and mapping for WFP in Somalia.

But Mr. Renk stressed that using AI to help make difficult decisions doesn’t compensate for the fact that most Somalis who need help will not be able to get it.

“It helps us to prioritise with stronger evidence,” Mr. Renk said. “But in itself, it doesn’t make the shortage of support acceptable.”