HARNESSING MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE FOR EAR-
LY FRAUD DETECTION AMONG BANKS IN HARARE, ZIMBABWE: INTERNAL AUDITORS’ PERSPECTIVE

HARNESSING MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE FOR EAR-
LY FRAUD DETECTION AMONG BANKS IN HARARE, ZIMBABWE: INTERNAL AUDITORS’ PERSPECTIVE

CHINGWARO LLOYD

MUCHOWE REGIS MISHEAL

NJAYA TAVONGA

The study explores the transformative power of utilizing machine learning and artificial intelligence
for early fraud detection among banks in Harare, Zimbabwe using qualitative research. Data
were collected through document reviews and in-depth interviews with bank internal auditors
and senior management. The study addressed three key research questions, namely, examining
internal auditors’ understanding of machine learning and artificial intelligence tools/systems for
fraud detection; understanding internal auditors’ perceptions on the effectiveness of machine
learning and artificial intelligence-based fraud detection systems; and identifying major challenges
faced by internal auditors during implementation of artificial intelligence fraud detection systems.
Internal auditors’ perceptions were gathered through in-depth interviews which were conducted
face to face and online. Findings from the study demonstrated strong consensus among internal
auditors on the potential power of machine learning and artificial intelligence in detecting fraud at
an early stage. In addition, the study revealed the potential benefits of utilizing machine learning
algorithms and artificial intelligence which includes enhanced speed in identifying anomalies,
improved accuracy, and the ability to detect fraud early, thereby enabling management to come
up with internal control mechanisms which can prevent fraud. Successful implementation of
machine learning and artificial intelligence-powered fraud detection systems require adequate
training and support from the organization’s leadership, and ethical considerations.

MET Mangement Review - MMR

2024