AI is an accelerator, not replacement for humans: WFP's Magan Naidoo

AI is helping humanitarian agencies move faster, cut costs and reach more people - but cannot solve hunger on its own, says Magan Naidoo, chief data officer of the United Nations World Food Programme

Artificial intelligence (AI) has become critical for aid agencies such as the World Food Programme (WFP) seeking to do more with less as humanitarian crises deepen and funding shrinks.

 

“But when it comes to addressing the core issues related to hunger, AI at best can act as an enabler and accelerator, helping get food to the right place at the right time, at lower cost and with greater precision,” said Magan Naidoo, WFP’s chief data officer, in a video interview with Business Standard.

 

“For example, one of our (WFP) solutions called enterprise deduplication uses photo data to match beneficiaries to remove duplicate data from the system. Now, you can imagine WFP has millions of records. We do have human teams going through them, which takes an immense amount of time. And it’s still prone to errors and cost. Now, we have AI that can crunch that data at high speed, especially if you leverage a supercomputer,” said Naidoo, who was in India for the AI Impact Summit. He was part of a panel discussion participated by Rome-based UN Agencies like IFAD, FAO and WFP at the Summit.

 

 

Optimising supply chains

 

WFP is using AI to optimise supply chain routes, ensure food reaches vulnerable people quickly and cost-effectively. Advanced modelling tools are being used to redesign logistics networks so that limited funds are used most effectively.

 

AI is also being used to strengthen food resilience at the grassroots. In several projects, machine learning models analyse soil samples, crop conditions and weather data to support smallholder farmers. In parts of Africa, AI systems advise farmers on irrigation — down to how much water to use and when — a critical intervention in parched regions.

 

“These are not grand, abstract projects,” Naidoo said. “They go down to the smallholder farmer, helping improve yields and build resilience.”

 

In floods and conflicts, AI-enabled satellite and drone imagery is helping WFP teams assess damage and identify stranded populations.

 

Digital rapid-assessment systems process images using machine learning to support search and rescue operations.

 

AI models analyse satellite data to assess buildings or map crisis-affected areas, allowing emergency teams to be deployed precisely and efficiently.

 

“These tools allow us to respond in a much more targeted way,” Naidoo explained. “Instead of broad deployment, we can send teams exactly where they are needed.”

 

WFP’s HungerMap LIVE and forecasting tools further use real-time monitoring and predictive analytics to anticipate food crises, enabling proactive rather than reactive interventions.

 

The growing reliance on AI comes at a time of financial strain. Humanitarian agencies are grappling with shrinking budgets even as needs rise. WFP requires an estimated $13 billion to assist 110 million people in 2026, yet faces funding shortfalls that have forced ration reductions in some operations.

 

Naidoo said that to address this, WFP is using AI internally to drive operational efficiencies.

 

In its communications division, for example, AI bots trained with established editorial guardrails can draft press releases in minutes — a task that previously took days. Human experts remain in the loop to review and approve outputs, but the technology significantly boosts capacity without additional hiring.

 

“We refer to this as human augmentation, not human replacement,” Naidoo said. “The bots don’t replace expertise — they scale it.”

 

However, Naidoo said that despite these advances, deploying AI in humanitarian settings is fraught with obstacles.

 

Only about 3.5 per cent of humanitarian staff possess expert-level AI knowledge, and many organisations offer minimal formal training.

 

This gap has led to instances of “shadow AI”, in which individuals experiment with tools without institutional safeguards.

 

Data quality poses another challenge. AI models require vast, reliable datasets, yet humanitarian contexts often lack robust data collection systems, particularly in hard-to-reach or conflict-affected areas. Infrastructure constraints — limited connectivity, electricity shortages and restricted access to high-performance computing — further complicate deployment, especially in the Global South where needs are greatest. 

FCI signs five-year deal with WFP

 

The Food Corporation of India (FCI) on Wednesday signed an agreement with the WFP to supply 200,000 tonnes of rice (upto 25 per cent broken) to support global humanitarian operations and combating hunger. The agreement is valid for five years from the date of signing and may be extended by mutual consent at the current price fixed at ₹2,800 per quintal which is valid upto March 31, 2026. It will thereafter revised annually on a mutually agreed basis.