Nigerian fintech company Lendsqr is developing an innovative artificial intelligence (AI) model that uses borrowers’ facial expressions and voice patterns to assess their eligibility for small loans—a potential game-changer in credit access for the underbanked.
The AI tool, currently reporting 76% accuracy, is designed to support Nigerians who lack formal financial documentation. It enables them to apply for microloans ranging from ₦30,000 to ₦50,000 (approximately GH₵264.85 – GH₵441.41) by simply speaking to a digital interface.
“Can we help vulnerable people prove their capacity and character, not through paperwork but through their words?” said Adedeji Olowe, CEO of Lendsqr. “That’s the thinking behind this AI project.”
Rethinking the Five Cs of Lending
Traditional lending typically relies on the five Cs—Character, Capacity, Capital, Collateral, and Conditions—to evaluate creditworthiness. However, Lendsqr’s AI model aims to simplify this process by focusing on just two: capacity and intention to repay.
Instead of filling out lengthy forms, applicants can interact with the AI via voice or video, responding to basic questions about their employment and how they plan to repay the loan. The model analyzes vocal tone, facial expressions, and speech content to predict repayment likelihood.
Testing and Transparency
Lendsqr is currently piloting the model using its own capital. The company plans to publish its research findings by Q3 2025, allowing other lenders—and even competitors—to adopt the model’s insights to power their own credit engines.
While Nigeria remains the primary market, the company also plans to pilot the model in Canada, where immigrants and international students often lack local credit histories and struggle to access credit.
“Africa is the priority because the need here is the greatest,” Olowe explained. “From Kenya to Ghana, Côte d’Ivoire to Malawi, and South Africa, the pattern is the same: the underbanked lack documentation and are locked out of credit systems.”
Transformative Potential
The stakes are high: only 6% of Nigerian adults currently access formal credit, and less than 12% of Nigeria’s 41 million small businesses can obtain financing—even as banks post record-breaking deposit figures.
Fintech lenders have attempted to bridge this gap, but many rely on expensive verification processes that inflate borrowing costs. Lendsqr believes its AI model can help change that equation for the better.
“Imagine lending to 10,000 people and 9,000 repay because your screening is better,” said Olowe. “That’s a major boost to your sustainability and profitability.”
What Comes Next
Lendsqr’s AI model is being partially funded by the Nigerian Ministry of Communications, Innovation & Digital Economy, in partnership with Google. The company intends to release the tool to the public once its accuracy reaches 90%.
“If successful, this won’t replace traditional lending for mortgages or car loans,” Olowe clarified. “But it could unlock access to foundational, life-changing credit—small amounts that can make a big difference.”
