As donor-funded agricultural data systems shrink, a Chicago Council on Global Affairs white paper argues that agribusinesses, insurers and lenders must step in, and that farmers who supply the data deserve a share of the value it creates.
DAR ES SALAAM, 29 September 2026 — The systems that tell governments, traders and farmers what is being planted, what prices are doing and where hunger is building are quietly weakening. A new white paper from the Chicago Council on Global Affairs says the private sector will have to help rebuild them. It will only do so, the authors warn, if the business case is clear.
The paper, From Data to Decisions: Understanding the Business Case for Private Sector Engagement in Food Security Data, was published this month. Its authors are Peggy Tsai Yih, Senior Advisor for Global Food and Agriculture at the Council, and Catherine Bertini, a Distinguished Fellow at the Council and former Executive Director of the World Food Programme. It draws on a series of dialogues that RF Catalytic Capital, the public charity affiliate of The Rockefeller Foundation, convened this summer. It also draws on an August 2026 meeting in Chicago, where 19 leaders from agribusiness, food companies, technology, philanthropy and research met to bring industry into the conversation.
A data gap opening as aid declines
The starting point is uncomfortable. Food systems face mounting shocks from climate change, conflict, tariffs and geopolitics. At the same time, many public and donor-supported systems that track production, weather, markets and food security are being cut back or closed entirely. The paper cites analysis of the damage done to humanitarian data by recent United States aid cuts.
Policymakers are turning to private investment to fill the gap. The authors caution, however, that the case for companies to invest is weakest in exactly the places where the need is greatest: fragile, conflict-affected and climate-vulnerable settings. Private firms facing rising risk may pull back from the regions where early warning and better data would do the most good.
The authors do not treat this as purely bad news. They describe the shrinking aid landscape as a chance to rethink how governments, philanthropy, sovereign wealth funds and private capital can work together as genuine partners. In that model, scarcer public money would be used strategically to create the conditions for markets to function.
Start with the market, not the dataset
A central argument is that data investment must follow market supply and demand. The authors observe that farmers respond to market signals. Farmers will not make long-term investments without viable demand, particularly where land is contested between food, fibre and fuel.
That means looking across the whole value chain, from farmgate to consumer. It also means broadening the focus beyond staple crops. The paper calls for monitoring of nutrient-dense, high-value crops such as fruits, vegetables, legumes, nuts, coffee and cocoa. It argues these crops improve diets, raise farm incomes and expand market opportunities. It also identifies shifting local consumer demand as an under-explored driver of food-system change.
National statistical agencies remain essential public infrastructure, the authors say. However, the agencies’ figures need to be timely, granular and useful to farmers, extension officers and advisory services. The paper also says valuable private data on prices, supply chains and production should be assessed for possible public-good use, while legitimate intellectual property is protected.
From early warning to early action
The paper calls for a much broader definition of agricultural risk. It lists pests and disease, extreme heat, rainfall variability, water stress, conflict, export bans, fertiliser and energy costs, and supply-chain disruption. It argues that historical data alone is no longer a reliable guide. It also says the most important warning signs are often political, market and social signals that satellites and drones do not capture.
In the authors’ view, ground-truthing is indispensable. Farmer observations, extension networks, field sensors and pest traps provide the hyperlocal picture needed to validate models and train artificial intelligence. That information must also trigger action on farm practices, credit, insurance, procurement and public policy. Better data, the paper notes, can make hard-to-price agricultural risks legible to underwriters, reinsurers and investors, and so make them insurable and financeable.
The paper also raises a governance warning. It notes that public sentiment and predictive signals can be manipulated. As AI and prediction markets become more influential, safeguards against misinformation will be needed.
AI that speaks the farmer’s language
For East African readers, one section stands out. The authors argue that AI will only deliver value if models are trained for local languages, dialects and farming realities. That includes women’s voices, local crop varieties, local names for pests and chemicals, and the imperfect photographs farmers take in the field.
For a region where much agricultural advice is delivered in Kiswahili and in local languages, the point is practical rather than academic. A pest-identification tool that cannot understand how a farmer in Iringa or Nyahururu describes a problem is of little use to that farmer.
Sharing without giving everything away
On collaboration, the paper separates pre-competitive activities from competitive ones. Pre-competitive work includes shared standards, common identifiers and joint infrastructure, which companies can support without surrendering commercial advantage. Competitive assets remain proprietary. The authors point to data commons and shared infrastructure that allow information to be pooled without opening up all underlying data.
They also describe a food-system “digital twin”. This would combine market intelligence, satellite imagery, weather forecasts and supply-chain data into a simulation that anticipates disruption and tests responses before a crisis arrives.
Farmers must share in the value
The paper’s sharpest argument concerns fairness. Farmers generate much of the ground-truth data behind early-warning systems, insurance products and market intelligence. The authors argue that farmers should therefore capture a fair share of that value. The paper suggests direct payments, data-sharing incentives, cooperative ownership or revenue-sharing as possible routes.
Because most smallholders cannot pay for information services themselves, the authors suggest that downstream actors fund those services. These include insurers, lenders, input suppliers, processors and governments, working through value-chain finance or contract farming. To justify the investment, the returns must be visible in lower insurance losses, better credit decisions, higher yields or more efficient supply chains.
Cooperatives feature prominently in the paper. The authors argue that aggregating data through cooperatives and other trusted bodies strengthens farmers’ bargaining power. That argument has obvious resonance for Tanzania’s AMCOS and for cooperative structures across the region. The paper insists, however, that trust matters as much as money. Farmers need to know what is collected, who uses it and whether they keep control. If they see a system as extracting value from their communities, financial incentives will not win them over.
Blended finance as the bridge
In high-risk and immature markets, the authors see blended finance, insurance pools and cooperative models as the bridge between public investment and commercial viability. Public or philanthropic capital would absorb early-stage risk, build common infrastructure and prove the model. The stated goal is not permanent subsidy but a pathway for private capital to enter and eventually sustain these systems.
Four recommendations
The paper closes with four recommendations:
- Build demand-driven data systems. Start from real market and value-chain needs, and extend monitoring beyond staple crops.
- Move from data to anticipatory action. Plan for uncertainty as the norm, and invest in local early-warning indicators.
- Create shared data infrastructure and trusted partnerships. Develop digital public infrastructure that supports trusted AI while protecting ownership.
- Align incentives and financing. Treat data as an asset with clear stewardship, and ensure farmers can benefit from the value it creates.
Why it matters here
For agricultural growth corridors, cooperatives, agro-dealers and agritech start-ups across East Africa, the paper offers both a challenge and a framework. The era of donor-funded data as a free public service is ending. Whether the systems that replace it serve smallholders or simply extract from them will depend on who invests, who governs, and who is paid.
The full paper is available at globalaffairs.org.