Nigerian financial institutions are turning to AI-powered accounting systems to fight fraud that rule-based auditors can no longer keep up with.
Researchers surveyed 186 professionals across Nigeria's banking, insurance, and FinTech sectors to measure how AI-enabled Accounting Information Systems affect fraud outcomes. Using multiple regression analysis, they found these systems meaningfully improve fraud prevention, detection, data analysis, and investigative effectiveness. They also tested whether Natural Language Processing played a role — and it did: NLP strengthened the relationship between AI systems and auditing quality by improving how those systems interpret language and explain their reasoning.
The Nigeria angle matters more than it might seem. Emerging markets often absorb fintech at speed without the legacy compliance infrastructure that slower-moving Western banks built up over decades. That gap makes AI auditing tools both more necessary and harder to validate — which is exactly what this kind of primary survey research tries to address. A sample of 186 professionals is modest, but it is primary data from practitioners inside the institutions, not vendor benchmarks.
The study leans on the Fraud Diamond Theory and the Technology Acceptance Model as its theoretical anchors — frameworks that are well-established but also well-worn. That the findings confirm what most AI vendors already claim about their products does not make them wrong, but it does mean the interesting work is in the replication, not the headline.