Credit risk assessment has emerged as the leading application of artificial intelligence (AI) among Kenyan lenders, with more institutions turning to the technology to streamline operations, ushering in a shift to data-driven lending that could speed up loan approvals and help curb default rates.
A new survey by the Central Bank of Kenya (CBK) on AI use in the banking sector shows that, while 50 percent of lenders are yet to adopt the technology, 65 percent of those already using it apply AI for credit risk scoring.
Adoption is highest among digital credit providers (DCPs), 80 percent of whom use AI to assess borrower risk. They are followed by microfinance banks at 75 percent, while commercial banks lag, with only 45 percent deploying AI for this purpose.
Credit risk assessment involves determining a borrower’s creditworthiness by assigning them a risk score or estimating their probability of defaulting.
This process determines how much interest a borrower will pay, and under the risk-based pricing model, it can take days and remains susceptible to errors—often resulting in delayed approvals or overpriced credit.
With AI, computer systems trained on data to mimic human intelligence and generate predictions, banks say they are making faster and more accurate decisions on whom to lend to and at what rate.
“The adoption of AI has brought about operational efficiencies, particularly in credit risk management, cybersecurity, and customer service,” the CBK said in a report on the survey published on Friday.
Cybersecurity is the second most common application of AI in the financial sector, with 54 percent of adopters using it for this purpose. Customer service follows at 43 percent.
Other common uses include fraud risk management (43 percent), electronic Know-Your-Customer (KYC) and onboarding (41 percent), general risk management (33 percent), product personalisation (32 percent) and anti-money laundering enforcement (32 percent).
Even among the 50 percent of lenders who say they have not yet adopted AI, uptake is expected to rise, with the majority already expressing interest or planning implementation.
Credit scoring is the most preferred application among them. According to the survey, 92 percent of commercial banks that do not yet use the technology plan to apply it in credit assessment, followed by 86 percent of microfinance banks and 77 percent of digital lenders. Overall, 83 percent of respondents said they intend to use AI for credit risk scoring in the future.
The trend comes as lenders face record levels of loan defaults, which hit 17.7 percent, with commercial banks alone holding Sh717 billion in non-performing loans—the highest on record.
Last year, CBK revealed that some lenders had begun using AI to monitor staff behaviour as part of efforts to combat employee-aided fraud—the first public confirmation that local banks were actively applying the emerging technology internally.
Globally, AI adoption in finance has continued to rise, though many international lenders primarily use the technology for fraud detection and anti-money laundering, unlike Kenya, where credit assessment leads.
