Malawi deploys WhatsApp AI to advise smallholder farmers who lack smartphones

MULANJE, Malawi — A WhatsApp-based generative AI chatbot backed by the Malawian government is being used to give farming advice to thousands of smallholder farmers who lack smartphones or reliable internet, officials and users say.

The tool, called Ulangizi — Chichewa for “advisor” — helped 59-year-old Alex Maere change crops after Cyclone Freddy stripped fertile soil from his farm in 2023. Maere said the chatbot recommended planting potatoes alongside corn and cassava; he followed the guidance, harvested a half-soccer-field plot of potatoes and sold more than $800 worth, enough to pay his children’s school fees.

Opportunity International, the nonprofit that developed Ulangizi, and Malawi’s agriculture ministry say the chatbot is intended to reach citizens in a country where more than 80% of the 21 million population depend on agriculture. The tool operates in Chichewa and English, accepts typed or spoken questions, and can reply with audio or text; it also allows users to upload photos of crops for identification.

Because many rural residents do not own smartphones or cannot read, the program pairs the AI with local "farmer support agents" who carry a device and bring the service to village groups. In the area around Mount Mulanje, 33-year-old Patrick Napanja visits clusters of 150 to 200 farmers, using the app to answer questions that previously frustrated extension workers. Richard Chongo, Opportunity International’s country director for Malawi, described that human role as the "human in the loop."

Officials and developers say the government supported the project in part because Malawi has recently faced a series of climate shocks, including cyclones and an El Niño-driven drought, that have worsened a food crisis and elevated agriculture as a central issue ahead of national elections. The U.N.'s International Fund for Agricultural Development estimates 33 million to 50 million smallholder farms across sub-Saharan Africa produce up to 70% to 80% of the region’s food supply, underscoring the potential reach of technology pilots like Ulangizi.

Private investment in agriculture technology for sub-Saharan Africa has risen sharply, from about $10 million in 2014 to roughly $600 million in 2022, according to the World Bank. Proponents say AI can help farmers identify crop diseases, forecast droughts, design fertilizers to boost yields and locate affordable machinery.

But developers and technology specialists caution that several barriers limit the technology’s reach and reliability. Malawi’s rural areas often lack steady electricity and internet service, and the country has many languages and widespread low literacy. Agents such as Napanja say village meetings can be dominated by waiting for responses to load; they sometimes hike to hilltops to find a signal.

“One of the biggest challenges to sustainable AI use in African agriculture is accessibility,” said Daniel Mvalo, a Malawian technology specialist. “Many tools fail to account for language diversity, low literacy and poor digital infrastructure.”

Mvalo also warned about the consequences of inaccurate AI outputs. For farmers operating on narrow margins, a wrong diagnosis of a crop disease or unsuitable treatment recommendation could destroy a season’s yield and undermine trust in the system. “Trust in AI is fragile,” he said. “If it fails even once, many farmers may never try it again.”

Opportunity International and Malawi’s agriculture ministry have aimed to limit such risks by programming Ulangizi to align with the ministry’s official extension guidance, Webster Jassi, the ministry’s agriculture extension methodologies officer, said. Jassi noted the service also leverages communal sharing: farmers who access the app help neighbors who cannot, spreading recognized practices through traditional social networks.

Still, scaling the program faces financial and technical constraints. Households need not only devices but also the means to afford data and charging. Extension agents can reach only a portion of the rural population, and training local staff to recognize AI errors and verify recommendations remains an ongoing task.

Developers and officials framed Ulangizi as one component of a broader strategy to modernize agriculture under challenging conditions. They said combining AI tools with human intermediaries, local language support and alignment with government advice increases the chance that technology will be useful and trusted.

As climate impacts and demographic pressures intensify across sub-Saharan Africa, Malawi’s pilot highlights both the promise and the limitations of applying generative AI to frontline food production. The program’s leaders say measured expansion, continued technical oversight and investment in rural connectivity will determine whether the approach can move beyond pilot success stories to durable, population-wide improvements in productivity and resilience.