AI levels the field: Kenyan farmers get smarter access to credit
For many smallholder farmers in Kenya, access to credit isn’t just a hurdle—it’s a dealbreaker. No money means no seeds, no livestock feed, and no way to weather the growing risks of climate change. Traditional banks are often reluctant to lend, viewing farmers as too risky, too informal. But now, a new tool powered by artificial intelligence is flipping the script.
The ‘Farmers Credit Scoring Model’, developed by Kenyan tech firm Pathways Technologies LTD and supported by the European Union and the German government through the Data Governance in Africa Initiative, promises faster, fairer, and smarter lending decisions.
Closing the credit gap
“This model changes the way we work from the ground up—it’s faster and more reliable and saves us a lot of work,” explains Henry Wachira Nyaga, Head of Credit of Fortune Sacco. With a total of 124,000 members, the agricultural cooperative is one of the largest in the country. With its help, the AI model was tested and trained in Kirinyaga county.
The stakes are high: Smallholder farmers account for almost 80 percent of agricultural yields and 70 percent of Kenya’s food production. Yet fewer than one in five have access to formal credit. Traditional banks often turn down loan requests because many farmers have no conventional credit history and therefore cannot offer sufficient collateral from the bank’s point of view. And the risks in agriculture are high: crop failures due to floods or droughts in particular have increased in recent years and are causing problems for the sector.
As a result, farmers are increasingly being excluded from the banks which poses a threat to their livelihoods. We have been working on this topic for some time. In GIZ, we have found a partner who believes in the idea which is great.
Joel Onditi, CEO of Pathways Technologies LTD
The IT company is based in Kenya and developed the model. GIZ supported the development of the model through the Data Economy Initiative, under the Data Governance in Africa Initiative and the Digital Transformation Center Kenya.
What makes this model different?
Unlike traditional credit models that primarily consult financial data and credit histories, this AI model taps into alternative data—crop yields, market sales, and payment behaviour—metrics that actually reflect a farmer’s potential. “We are all farmers ourselves and of course we know the challenges,” says Henry. Fortune Sacco has therefore been using alternative data sources for lending for some time, such as crop yields, sales figures and payment delays. This was previously collected and managed manually.
Pathways Technologies LTD has now completely digitised this data and developed an AI that will replace the entire process in the future: “The data was already there, it took us about four months to get the model ready for use,” says Joel.
The data records from KALRO (Kenya Agricultural & Livestock Research Organization) were also crucial. The authority has data sets from 6 million farmers in Kenya, which they have made available for this project.
We no longer have to do the calculations ourselves but can simply feed the data into the tool and trust that it will tell us whether we can grant a loan and how much
Henry Wachira Nyaga, Head of Credit of Fortune Sacco
The more data, the more accurate the AI
“The main work has been done, namely setting up the model,” says Joel. This model can now also be made available to other cooperatives and trained based on their data sets. According to forecasts, more than 20,000 Kenyan smallholder farmers could benefit from better and faster access to financial services.
In the future, it could also be extended to other African countries that want to strengthen their farmers with technological financial solutions. GIZ is currently examining this step on behalf of the Data Governance in Africa Initiative—especially neighbouring Tanzania and Rwanda hold great potential.
For thousands of African farmers, a digital tool born in rural Kenya may soon mean the difference between subsistence and sustainability—and between being overlooked and being seen.
