Collateral Not Required: How Kifiya’s AI Platform Signals a Path to Rewriting Ethiopia’s Financial Innovation Story

Ethiopia’s financial sector has remained shallow, measured across several parameters. Nearly half of the adults in the country still don’t have an account at a formal financial institution, while almost 31% of all bank branches in the country are located in the capital. For much of the rural population and low-income households, credit, investment instruments, and formal saving institutions remain out of reach.

Digital financial services once carried the promise of transformation. Banks introduced ATMs 24 years ago, followed by Point of Sale (PoS) machines a decade later. The launch of licensed mobile money services four years ago, led by Telebirr, was also poised to be a major turning point. Many within Ethiopia’s financial sector had predicted a financial inclusion success story like that of M-Pesa’s in Kenya, with over 75% of the populace using mobile money services. However, as the latest World Bank Findex database would go on to show, adult account ownership had increased by just 3% by 2024 across Ethiopia, even in the post-Telebirr era. 

Yet beneath the slow pace of account growth, one metric of financial maturity has shown the potential of transforming lives: digital credit. Around 4.1 million new customers accessed digital credit just in the last year, raising the total number of credit accounts in the country to around 10.5 million. 

At the heart of this surge lies the expanding role of artificial intelligence (AI), with finance being one of the high-potential AI use cases identified by recent sector studies. Once confined to niche tech startups and government institutions such as the Ethiopian Artificial Intelligence Institute, AI is now reshaping financial innovation across Ethiopia in concrete ways. 

In January of 2022, Kifya Financial Technologies partnered with the Cooperative Bank of Oromia (COOP) to launch a digital lending product dubbed Michu. Leveraging Qena, an AI-powered digital lending platform by Kifiya, the bank was able to provide uncollateralized loans targeting Micro, Small, and medium businesses (MSMEs), a segment historically shut out of formal credit markets due to the lack of collateral.

The partnership, anchored in an intelligent credit scoring system, would immediately be embraced by Ethiopian MSMEs. By August of this year, Michu would go on to disburse over 30 billion Birr to more than 2 million digital loan accounts. Even more impactfully, through a specialized product for women dubbed Michu Kiyya, the platform disbursed over 9.3 billion Birr to over 1.3 million women, tackling the crippling gender disparity that has long undercut financial inclusion success in Ethiopia. The Qena platform and its data-flow design are singled out in research as an instructive example of alternative credit scoring in action.

The ripple effect was unprecedentedly swift. Five additional banks adopted Qena to offer collateral-free credit. Efoyta (Wegagen Bank), Abol (Bunna Bank), Malefiya (Enat Bank), Lewedaje (Amhara Bank), and a Sharia-compliant embedded finance service dubbed Ansar with Zamzam Bank. Collectively, the total amount of collateral-free disbursements would reach close to 36 billion Birr, bolstered by the Sustainable Access to Finance to Enable Entrepreneurship (SAFEE) program, co-created with the Mastercard Foundation.

 Kifiya’s management is far from satisfied with its recent success. During the 3^rd knowledge series last month, hosted by the company in partnership with the Mastercard Foundation, CEO Munir Duri urged stakeholders to innovate in creating new markets rather than merely competing in existing ones. A sentiment echoed in Kifiya’s recent moves. The company has expanded into smart agriculture, micro-insurance, and contract farming, anchored in two technical pillars it calls Intelligent Data Decisions and Intelligent Financial Services.

These “meta-models” ingest sector-specific data, soil and weather for smallholder farmers, mobile-money and PoS activity for urban traders, and apply machine-learning methods (from logistic regression and random forests to neural networks) to estimate repayment capacity and recommend tailored products. For an urban trader, inputs can include mobile-money activity, point-of-sale transactions, and nearby purchasing trends.

The ingenuity also extends in powerful ways for inclusion. Kifiya’s Origination and Verification Platform (OVP) bridges online and offline environments: field agents can collect data, validate IDs, and complete ‘know your customer’ procedures without stable connectivity. That design has helped roughly 130,000 smallholder farmers access over 730 million Birr in input credit and given many MSMEs their first financial footprint. Such approaches echo a broader Global System for Mobile Communications Association (GSMA) finding that mobile-enabled, offline-capable tools and edge computing are essential for extending AI benefits beyond well-connected urban pockets. 

But the GSMA analysis also flags persistent constraints that matter for scaling these innovations: limited access to diverse, sector-specific local data (especially datasets in Ethiopia’s many languages), infrastructure gaps, and weak incentives for private firms to share the third-party data that alternative credit scoring often requires. Without richer datasets and clearer data-sharing frameworks, AI systems risk entrenching bias or excluding those with limited digital footprints. 

The report recommends a package of enablers that speak directly to the next phase of Ethiopia’s financial AI evolution: strengthen the national data ecosystem, invest in computing and local cloud/HPC capacity, adopt financing models to de-risk AI investments, and build talent through public-private partnerships and research hubs. It also calls for tailored strategies to reach rural and low-connectivity users, leveraging SMS, USSD, and other low-bandwidth channels while ensuring that model development is rooted in local linguistic and contextual data. These priorities mirror the practical work Kifiya and its partners are doing, while outlining the systemic investments still required to move from successful pilots to sustainable, inclusive scale. 

AI has emerged as a quiet but transformative force in Ethiopia’s financial sector. With a recent national AI policy, improving power and connectivity, and demonstrable use cases such as Qena, the country is experimenting with how intelligent systems can widen access and tailor products to previously invisible customers. Yet the path ahead is not automatic: closing data gaps, building compute and human capital, and creating incentives for responsible data sharing will determine whether these early gains become lasting inclusion. Kifiya has shown what is possible; now the challenge is for more players, policymakers, and innovators to match that ambition and take financial AI to scale.