[X]

We make enough food, so why are South Africans going hungry?

Artificial intelligence (AI) researchers in South Africa and the United Kingdom (UK) have developed a prototype tool that will help government and farmers solve one of South Africa’s biggest conundrums: we produce enough food, yet many of us go hungry every day. 

“Official reports show that as a country we are food secure at the national level, but at the household level we are food insecure – this is a major challenge,” says Dr Essa Suleman, a food system and animal health expert at the CSIR. 

He is part of a team of researchers from the CSIR who, together with the Centre for Environment, Fisheries and Aquaculture Science (Cefas) in the UK, are working on the One Food Risk Tool.  

The tool can sift through and analyse masses of public data around threats to food supply chains. Once the data is parsed, the tool can guide policymakers and decision-makers in government, as well as farmers and producers, to solve problems.  

“In poultry, for example, a big question for farmers is how to deal with pests and infectious diseases,” says Suleman. 

“We have regular outbreaks of avian influenza and other diseases, which means tens of thousands of chickens are culled. That’s a massive socioeconomic loss and can drive prices up for consumers.” 

Limpopo-based poultry farmer Mologadi Madisha agrees that avian influenza outbreaks are becoming trickier to navigate and says it can be described as “a likely collapse of operations for most small-scale farmers”.  


Limpopo-based poultry farmer Mologadi Madisha inspects the health of a chicken. She says avian influenza outbreaks are a threat and becoming trickier to navigate. The final version of the One Food Risk Tool, currently under development by researchers at the CSIR and in the UK, will feature several advanced AI capabilities to predict such potential disease outbreaks by analysing economic trends, climate conditions, disease surveillance reports and other relevant factors. It will then prescribe strategies to prevent losses and alleviate strains on the food supply. 

“Usually, small-scale farmers do not have access to early warnings,” she says, adding that she is hopeful that new technologies could alert her to outbreak risks, where to get help and generally how to control and mitigate disease outbreaks on her farm. 

Suleman says that the final version of the One Food Risk Tool will indeed have several advanced AI capabilities to predict potential disease outbreaks by analysing economic trends, climate conditions, disease surveillance reports and other relevant factors. Then, it will prescribe strategies to prevent losses and constrains on the food supply. 

But, he says, the tool will go even further up and down the value chain because food systems are incredibly complex.  

To illustrate part of this complexity, we can look at poultry again, says Suleman: chickens feed on maize, so maize farming is another critical link in that value chain. “We need to know if there is going to be another drought in Gauteng and how will it affect maize production in the country five years down the line,” says Suleman.  

“If we are going to have a crisis, can we stockpile maize in advance? Are there other countries we can import from? Can we get farmers elsewhere to produce more?” 

Suleman says the risk tool will help answer some of these questions. The tool forms part of the broader international One Food programme, which approaches the three interrelated pillars of health – human health, environmental health and animal health – from a food perspective.  

“The One Food programme puts food at the centre of those three pillars because, in essence, every organism requires food to survive,” he says.  

The CSIR is one of the lead institutions driving the programme in South Africa alongside Cefas, with Suleman as the technical co-lead for the broader One Food programme. 

“As an organisation, the CSIR has a diverse set of expertise – in climate, nutrition, food production, animal health and AI – so we are in a prime position to lead some of this work,” says Suleman. “We also have partnerships across the science, technology, and innovation and academic ecosystems in South Africa, and strong international networks.” 

He says the team developed the prototype in a year and they expect to hone its predictive capabilities within the next four years. 

The team presented the prototype at the Food and Agriculture Organization (FAO) of the United Nations in Rome, Italy, in February 2025, and is testing it in specific scenarios. It can currently assess risks using information provided by expert scientists, public databases and research reports. 

"We have a list of over 70 hazards, and if you try and score for each hazard, you have to generate nine individual scores for each health pillar,” he says.  

“The amount of computation is massive, especially when we start finding patterns in historical data to make predictions and prescribe solutions. Eventually it’s going to be beyond the realm of human capability, and that’s where AI comes in.” 

Suleman says that in his view, the risk tool will be one of very few AI products being developed globally that will make a significant, tangible impact on people’s lives. 

“The ultimate vision is that this tool will be accessible to other countries to plug in their own data and produce localised suggestions for interventions,” he says. “Over time, one can then continuously monitor whether these interventions are working and that will be the real benefit in the long-term.” 

He says they expect government representatives, policymakers and scientists to use the tool to address the imbalance between sufficient food production and food insecurity. 

“The tool will help all of us bring to light things we may not have thought of using, including traditional approaches, but which may actually have significant impact on the food system,” he says. 

“As a country, for instance, we have several diseases affecting livestock production, but we do not know the impacts of those animal diseases on things like biodiversity; or how climate change in turn affects animal diseases; or even how all of these factors together affect people’s access to enough nutritious food.” 

Even wildlife tourism can be negatively impacted by risks to the food system, he says. 

For example, an outbreak of foot-and-mouth disease on a cattle farm next to the Kruger National Park can spread to the wild buffalo roaming there. Conversely, buffalo are reservoirs of pathogens like the foot-and-mouth-disease virus.  

“So, One Food is not just about food production and food sustainability, but also about how it impacts other life on land and human social interactions.” 

While the CSIR has had significant support from the Department of Science, Innovation and Technology for the broader One Food programme and to develop the risk tool prototype, Suleman says more buy-in is needed from the international community and other national departments overseeing agriculture and the environment.  

“We have a five-year road map for the development of the tool but that is dependent on the funding we can secure.” 

Besides becoming a vital decision-support tool and scientific research instrument, Suleman says plans are in place for it to one day be in the pockets of farmers themselves, as a mobile application providing alerts and suggesting actions to mitigate climate, disease and other risks to their day-to-day operations. 

The One Food Risk Tool forms part of an international One Food collaboration between South Africa, the United Kingdom and more than 20 global partners. It addresses SDG 2, SDG 3 and SDG 15. 

Contact Person

Essa Suleman

esuleman@csir.co.za