A Business and Data Analyst, Olatoye Agboola has called on banks, fintech firms and digital lenders to embrace cross-disciplinary collaboration to tackle bias in Artificial Intelligence (AI) and machine learning (ML)-based credit scoring systems.
Agboola, a New Jersey City University-trained IT expert, made the call amid growing concerns that automated lending tools, while expanding access to credit, risk entrenching unfairness when left unchecked.
Agboola, in a study titled: “Auditing Bias in AI and Machine Learning-Based Credit Algorithms: A Data Science Perspective on Fairness and Ethics in FinTech”, called for a multi-stakeholder approach involving data scientists, ethicists, regulators, policymakers, and affected communities.
He stressed that fintechs must work closely with data scientists, ethicists, regulators, policy-makers, and affected communities to ensure fairness is embedded in their systems.
In his view, increased collaboration with regulators would foster a more inclusive financial ecosystem where access to credit is determined by merit and responsibility rather than biased algorithms that replicate past inequalities.
Agboola argued that algorithmic bias is not just a technical challenge but also an ethical, legal, and societal one.
The tech expert noted that strong governance structures, clear regulatory guidelines, and a culture of transparency are essential for holding financial institutions accountable.
According to him, without these measures, experts warned, AI systems risk reflecting and reinforcing long-standing social inequities.
“FinTech firms should foster collaboration among data scientists, ethicists, legal experts, and credit policy professionals. Establishing internal AI ethics committees or review boards can provide oversight and ensure that fairness considerations are prioritised alongside business objectives,” he said.
Agboola also made case for continuous innovation in auditing methods, the adoption of fair-by-design principles in AI model development, and sustained dialogue between fintechs and regulatory bodies such as the Central Bank of Nigeria (CBN).
He also noted the need for new fairness metrics, practical applications of explainable AI in lending, and tools to identify and correct unintended bias over time.
“Building trustworthy AI credit systems is not only a technical task but an ethical imperative. Financial institutions must prioritize fairness as a core design principle rather than an afterthought,” the study stated.
In another study titled: “Predicting Loan Defaults Using Ensemble Machine Learning and AI-Driven Credit Scoring Models: A Comparative Study”, Agboola noted that AI and ensemble machine learning models offer more reliable tools for predicting loan defaults than traditional credit scoring methods, opening new possibilities for lenders and fintech firms in Nigeria and other developing markets.
“AI credit models must not remain black boxes. Explainable mechanisms are necessary to align innovation with fairness and accountability.
“The implementation of explainable AI mechanisms guarantees such intricate models will be able to achieve some level of transparency in areas where the explanation of black-box algorithms has long been an issue of concern in regulated financial settings,” he said.
Discussion about this post