Beyond the harvest: AI accountability in Nigeria’s agricultural future

AI in Nigerian agriculture is being pitched as a lifeline at a time of climate shocks, inflation, and conflict. The World Food Programme projects that 33 million Nigerians will face food insecurity in 2025. Against this backdrop, agritech founders are racing to deploy AI-driven tools that promise to squeeze more output from less land and water.

Startups and established platforms pitching into the Nigerian agriculture frame AI as a multiplier of scarce inputs.

Yet, smallholder farmers, the bedrock of Nigeria’s agriculture, accounting for about 80% of the nation’s food according to FAO, sit at the sharp edge of a widening crisis.

Read also: Zoho reports 75% growth as Nigerian businesses lead global shift in responsible AI

Agritech startups in Nigeria

Kitovu, a Nigerian agritech named by several coverage pieces as part of the country’s AI agenda, markets data-driven advisory tools that use remote sensing and agronomic models to provide tailored soil and crop recommendations.

Coverage of Kitovu places it among a group of local innovators pushing precision advisory services to smallholders. 

As of 2021, Kitovu has reportedly enabled over 7,000 smallholder farmers to cut down input costs by 30%, increase their yields by 50%, and achieve 100% sales of their produce. And, as of 2023, the company reports that it has trained 307 agents who work with about 12,000 farmers in Oyo, Gombe, Niger, Jigawa, Kaduna, Kano, and Plateau states.

“Our journey started when we realised that farmers desperately needed precision agriculture but the solutions available weren’t designed for their realities,” said Emeka Nwachinemere, Kitovu’s founder. He recalls his first farm after youth service in 2014:

“Running that farm thrust me into the world of agriculture, showing me the problems faced by smallholder farmers firsthand. It was a real personal problem, and I decided I was going to fix it.”

Kitovu’s early IoT device, SoilSense, failed because of poor rural connectivity and affordability. “So we pivoted to using remote sensing,” he explained.

Today, we are leveraging AI and remote sensing to create hyper-local, crop-specific advisories tailored to Nigeria’s fragmented soils, erratic rainfall, and pest pressures. – Emeka Nwachinemere

“One agent manages 20 to 50 farmers and attends farmers’ engagement meetings to introduce Kitovu and onboard farmers.”

AgroScout (an Israeli firm, present in over 30 countries), an international precision-agriculture firm, advertises AI-powered image analysis and field-wide pest and disease monitoring down to leaf level, converting drone or smartphone images into actionable treatment reports. Its platform is built to flag pests early, a problem that can otherwise decimate yields in weeks. 

AgroScout claims to have produced over 10 million images, over 20 crops and over 400 acres across different countries. 

Hello Tractor operates at a different layer – mechanisation. The platform’s tractor-sharing model is now active in multiple African countries and has launched mechanisation hubs and services in Nigeria, promising farmers access to machinery without full ownership costs. 

Hello Tractor reports over two million smallholder farmers in its marketplace being serviced by over 5,000 farm equipment owners, boosting farmer incomes by 227%. It also reports over 2.5 million smallholder farms. 

“I saw a paradox – hardworking farmers producing far below potential because they lacked access to modern tools,” said Jehiel Oliver, CEO of Hello Tractor. “For example, there is a dearth of tractors in Nigeria, and the ones that do exist are often out of reach for smallholders—either owned by large commercial farmers or held within ineffective government programs.“

Hello Tractor closes that gap with IoT devices and AI-enabled software that connect tractor owners to farmers—like ride-hailing for agriculture. We analyse booking trends, weather, and soil data to ensure timely, affordable mechanisation. – Jehiel Oliver

Farmcrowdy – present in over 3 countries – one of Nigeria’s earliest digital agriculture platforms, has built scale around input financing, training and market access, claiming to have supported thousands of farmers across multiple states. Its model blends digital matching, sponsorship and agronomy support.

Farmcrowdy, which has raised over $1 million in funding, claims over 420,000 customers and over 2,000 commodities. 

Maigari Ibrahim, CEO of FarmEasy, who has been selected as a mentor for the Llama Impact Accelerator by Meta, an initiative focused on empowering startups across Sub-Saharan Africa to build real-world solutions using open-source AI with Llama, frames his company’s mission as stripping out friction for the smallest farms:

“Technology can bridge the knowledge and resource gap that keeps small-scale farmers from competing. But if the tools are built without understanding their realities—seasonal income, limited connectivity, and the way they share knowledge—then adoption will be slow, and the benefits won’t be equally shared.”

Taken together, these technologies promise efficiency, including better water management, earlier pest detection, improved matching of farm services and, ultimately, better yields.

The pitch is attractive at a time when multiple agencies warn of expanding food insecurity in Nigeria. And recent joint assessments and humanitarian agencies estimate tens of millions of Nigerians face high levels of food insecurity in 2024–25.

According to public material, many of these solutions are still at the pilot or early-adoption stage in Nigeria. Company pages, press notes and sector reporting indicate trials and demonstrations, but rarely provide full, independently verified impact assessments.

AgroScout and similar firms publish case studies describing accurate pest detection and field-level monitoring. These are useful demonstrations, but they do not always translate into easily replicable roll-outs across fragmented smallholder farms.

AgroScout’s product literature, for example, emphasises leaf-level AI analysis and automated reporting which are core technical capabilities that require reliable imagery, stable connectivity and farmer trust. 

Hello Tractor’s expansion rests on the logistics of pooling tractors, hub servicing and local operators. Its model can lower the capital barrier for mechanisation, but it also introduces dependencies on hub maintenance, fuel, spare parts and scheduling systems that must work across dispersed villages.

Hello Tractor’s recent announcements about new hubs in Nigeria show the model’s promise, but they also underline that mechanisation is at least as much a business-logistics problem as it is a technology one.

Kitovu and other local agritech companies promote remote sensing and advisory systems designed to reduce waste and fine-tune inputs. Press coverage and sector analyses cite these startups as emerging players, but public evidence of large-scale water-saving or yield-increase metrics in Nigeria is sparse.

Where companies announce trial gains (for example, reductions in water use or better input efficiency), the numbers are often preliminary and framed as pilot outcomes rather than generalisable results. Independent verification of such claims is limited in the public record. 

Who the tools reach and who they don’t

Technology’s value is judged by its reach to small-scale holders across diverse geographies. The country’s own agricultural profile makes this a pressing issue. FAO and IFAD reporting note that more than 70 per cent of Nigerians engage in agriculture at subsistence levels and that smallholder farmers produce the bulk of the nation’s food – 80%. 

“Instead of expecting farmers to adapt to tech, we adapted tech to them,” Nwachinemere said.

“Through our farmers’ service centres and over 1,000 trained field partners, farmers receive AI-powered advisories in their local languages, bundled with access to quality inputs, storage, and markets.“

This hybrid ‘high-tech, high-touch’ approach ensures even the most remote, low-literacy farmer can benefit without needing to own a smartphone. – Emeka Nwachinemere

These farmers are often resource-constrained, weather-dependent and remote. These are conditions that complicate high-tech interventions. 

“Farmers don’t need smartphones; they work through local booking agents, trusted community members who use our app on their behalf,” Oliver explained. 

AI optimises routes, predicts demand, and keeps tractors running, ensuring even the most remote farmers get service. On the supply side of the marketplace, we use a tremendous amount of data analytics to ensure farm equipment is optimised and services are tracked end-to-end. – Jehiel Oliver

Maigari notes that in some pilot projects, “The very farmers who could benefit most from yield-boosting tech are often the ones least able to pay for it upfront. That’s where design choices matter—things like offline functionality, shared devices, or pay-as-you-harvest models could be the difference between uptake and abandonment.”

At the same time, food-security forecasts underscore the stakes. International monitoring systems and humanitarian reports have flagged tens of millions of Nigerians facing acute or high levels of food insecurity in 2024–25, as mentioned earlier.

Any agritech that fails to account for cost, usability and distribution risks becomes an optional luxury for wealthier farmers while leaving the majority vulnerable. 

The practical barriers are straightforward. Connectivity, device costs, digital literacy, spare part availability and up-front service charges. If an AI tool requires frequent sensor maintenance, consistent satellite imagery or persistent data connectivity, it will struggle to scale among smallholders who farm less than a hectare and rely on seasonal cash flows.

The result can be a two-tier system in which the first adopters reap gains while the most vulnerable see little change.

Dolapo, a palm oil plantation farmer in Ile-Ife, Osun, echoed these barriers:

“Availability, cost, technical expertise amongst others,” he said when asked about challenges. “First, these tools should be made readily available to local farmers. The tools should be cost friendly to avoid cost discouragement.”

Accountability gaps: design, deployment and policy

The tech-first logic of many agritech ventures elevates scalability and automation. That can be positive, but it also risks sidelining accountability.

Three accountability shortfalls stand out:

Data and transparency. AI advisory systems feed on farmer data: soil tests, geolocation, yield histories and input purchases. Publicly available company material is generally thin on how data is stored, shared or monetised. Without clear data-use policies and localised consent processes, farmers may unknowingly surrender valuable information about land, yields or transactions.

This information could be used commercially without the farmer’s benefit.

Proof and procurement. Where pilot results are shared, they are often company-sourced. Independent, third-party evaluations and detailed trial protocols are scarce in the public record. Policymakers and funders need robust evidence of impact, not only manufacturer claims, to justify scaling and subsidy decisions.

Access and affordability. Mechanisation hubs, sensor networks and image-analysis platforms all carry ongoing costs. For smallholders to benefit, models must be designed with rural cash cycles in mind, including credit, pay-as-you-go models and local training. Otherwise, those tools remain optional add-ons for better-resourced producers.

“Our FarmEasy OS can run offline, in local languages, and is built to work without internet for days,” Maigari explained. “We design around low-resource conditions because if the tech fails when the network drops, it’s useless to most rural farmers.”

Meanwhile, “Our Pay-As-You-Go Tractor Finance lets booking agents become tractor owners through small payments,” Oliver said. “It’s turning local entrepreneurs into service providers, expanding access, and keeping value in rural communities. Our underwriting is based on data within their booking app, unlocking opportunities for traditionally unbanked, using data and AI.”

Maigari adds: “It’s about whether the tech actually fits into the rhythms of rural life. You can’t expect a farmer to risk a season’s earnings on an unproven system, and you can’t hold them responsible for failure if the infrastructure wasn’t there to make it work.”

On his own, Nwachinemere says, “Most agritech solutions focus on one piece of the puzzle: input delivery, advisory, or markets. Kitovu is different: we built an end-to-end ecosystem… From AI-powered advisory (YieldMax) to affordable input financing and storage (StorageX) to market linkages (eProcure), we support the farmer through the entire cycle.”

Nigeria’s National Agricultural and Innovation Policy (NATIP 2022 – 2027) and strategic and humanitarian reports have heightened awareness of food risk. 

The public sector is increasingly part of the discussion about how to deploy digital agriculture responsibly. Still, concrete regulatory frameworks for AI in agriculture, covering data protection, explainability of recommendations and farmer recourse, are only nascent in public discourse. The space is urgent: with more than 30 million people projected to face food insecurity in some projections, the cost of getting accountability wrong is national. 

Bottomline: AI in Nigerian Agriculture needs accountability before scale

AI-driven tools can improve water use, sharpen pest control and unlock tractor access for thousands of farms, but only if their deployment is deliberate and accountable. Nigeria’s agricultural future cannot be outsourced to algorithms without those algorithms being explainable, affordable and aligned with farmers’ seasonal realities.

Experts warn that without accountability, these tools risk deepening inequalities rather than alleviating them.

“Technology should make farming easier, not harder,” Maigari said. “If it’s just there to impress investors or tick an AI box, it will fail. The measure of success is whether farmers see real gains in yield, income, and access to markets.”

Oliver shared the story of Fatima, a farmer in Kaduna, who used to till by hand for three weeks—often missing the planting window. “With Hello Tractor, the work took a single day,” he said. “Her maize yield jumped 40%, she expanded her land, and now funds her children’s education.”

For farmers like Dolapo, the promise of agritech remains aspirational but powerful. “I believe they are quite miraculous. As in all aspects of life, technology usually helps ease the burden of work,” he said. “In agriculture, technology usage will go a long way to help secure food stability, and make farming more effective and efficient.”

The companies named here (FarmEasy, Kitovu, AgroScout, Hello Tractor and Farmcrowdy) are part of an ecosystem that could help avert a worsening hunger crisis. The equally important actors are regulators, funders and the farmers themselves.