Page 976
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
The Impact of Artificial Intelligence on Financial Inclusion in
Zimbabwe's Banking Sector: Challenges and Opportunities for
Expanding Access to Financial Service
Justin Tinashe Muswaburi
1
, Precious Mandava
2*
, Nyasha Mpondwe
3
, Maruva Mumanyi
4
2,3,4
Department of Business Management, Lupane State University, Box 170 Lupane, Zimbabwe,
1
First Banking Corporation (FBC), Zimbabwe
*
Corresponding Author
DOI: https://doi.org/10.51583/IJLTEMAS.2026.15020000087
Received: 25 February 2026; Accepted: 02 March 2026; Published: 19 March 2026
ABSTRACT
The purpose of the study was to investigate the effect of artificial intelligence (AI) on advancing financial
inclusion in Zimbabwean banks through AI credit scoring, chatbots, anomaly detection systems, and predictive
analytics. The researchers used a mixed-methods research design which included data collection from 293
respondents who completed the questionnaire and 12 participants who came for the interview. Artificial
intelligence (AI) technology enables loan processing by making financial products more accessible through its
three main functions which are usability and safety in transactions and financial literacy training. The researchers
found that organisations strongly supported AI systems for decision-making and fraud detection yet users still
had concerns about how AI credit assessments and chatbot operates. AI adoption process faces critical obstacles
which originate from four sources which are digital illiteracy, poor Internet access, excessive application costs
and the rural-to-urban divide. Anomaly detection systems had the most significant impact on financial outcomes
since they explain 62.3% of the outcome differences which AI technologies produce. The research found that AI
functions as an essential instrument for advancing financial inclusion in Zimbabwe through its ability to enhance
banking access and operational efficiency and secure banking services. The study recommended the
establishment of more accessible AI systems for decision-making, providing digital literacy programme
improvements through regulatory support, and creating special resources for communities that lack essential
services.
Keywords: Artificial Intelligence, Financial Inclusion, Credit Scoring, Chatbots, Predictive Analytics
INTRODUCTION
Artificial intelligence (AI) is establishing itself as a disruptive technology for international finance as it enables
automatic processes, data-based financial assessments and reduced operational expenses. Research showed that
AI credit scoring systems which use automated data sources rather than traditional credit records, create new
opportunities to provide loans to people who lack official credit documentation (Nuka and Ogunola, 2024). AI
banking services which include automated customer assistance, efficient on-boarding processes and instant fraud
detection technology, help financial institutions to enhance their operational efficiency while reducing expenses
and providing better service to under banked populations (Lee, 2024). Digital infrastructure improvements
enable these technologies to significantly decrease transaction and administrative expenses which currently
prevent low-income and rural people from accessing formal banking services.
Zimbabwean banks have experienced increased digital finance adoption throughout the last several years. This
created an environment which supports the use of AI to develop banking innovations. In 2025, Kawara,
Muchara, and van der Poll carried out a study to investigate financial inclusion among Zimbabwean commercial
banks. The results showed that financial inclusion had improved among individuals. The existing gaps between
rural areas and other marginalized populations demonstrate major problems. Masuka 2025, conducted a study
Page 977
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
on smallholder farmers and the results highlighted that these farmers could not access multiple financial services
due to financial illiteracy, education level, household characteristics and rural banking infrastructure. Digital
banking systems faced structural inequalities from literacy, income, geographical barriers, and service access
limitations which prevent more people from using them. AI-based banking solutions would create new
possibilities to help Zimbabwe's underserved populations through their services.
Adoption of Artificial intelligence require specific connectivity solutions, data infrastructure systems, regulatory
frameworks and social inclusion methods. Nuka and Ogunola, 2024 analysed the use of AI and machine learning
as tools for financial inclusion.
The researchers found out that whilst these concepts broaden access to credit and improve risk assessment, they
also harbour risks like algorithmic bias, lack of transparency, weak data governance, and data privacy-related
issues. Thu, Doan, and Quy, 2024 evaluated credit scoring fairness by analysing machine learning models which
have received fairness enhancements. Advanced machine learning models produced biased results due to the
need for auditing and correction. This demonstrated that automated credit evaluation perpetuated social
disparities without fairness constraints.
Regulation and governance functioned as dual components which requires equal assessment. Makore 2024
contend that financial inclusion should receive proper evaluation through analysis of regulatory frameworks
which include transparency requirements, consumer protection measures, data privacy regulations, and
accountability systems for automated decision-making.
Financial institutions faced difficulties with their AI transformation due to customers who lack digital skills and
minimal data presence experience adverse effects from the absence of required frameworks. People involved in
these situations faced impossible battle to prove their case against the algorithmic decision. This denied their
credit application since they lacked knowledge about the reasons and available options for appeal. The
governance issues became essential for Zimbabwe as the country faces two main problems which include
undefined regulatory frameworks and incomplete digital connectivity.
There is insufficient research evidence from 2023 onwards to determine whether AI-based banking services in
Zimbabwe actually provide financial access to underserved groups. The world has adopted digital finance as an
AI-based solution which Zimbabwe now implements to achieve financial inclusion through its digital financial
services. Research shows that AI technologies create new access barriers when their systems operate without
fairness mechanisms, transparency practices, and data management systems (Nuka and Ogunola, 2024; Thu,
Doan and Quy, 2024; Makore, 2024). This empirical gap requires country-specific studies to assess how AI
implementation in Zimbabwean banks impact financial access.
The research aims to investigate how AI adoption affects financial inclusion in Zimbabwe's banking sector. The
research first identified critical challenges and institutional barriers to AI implementation in banking. The study
used evidence to create ethical and inclusive AI adoption guidelines for the banking sector. The research
proceeded with the following objectives:
(1) To measure the association between exposure to AI-enabled banking services and indicators of financial
inclusion- account uptake, usage frequency, credit access among previously unserved clients.
(2) To identify infrastructural, institutional and data-governance barriers limiting AI-driven inclusion
(connectivity, data integrity, regulatory safeguards, digital literacy).
(3) To assess customer perspectives of trustworthiness and transparency and impartiality and recourse options
which AI banking services offer their users who belong to different user groups (i.e. urban, rural, women,
informal sector, etc. smallholder farming).
(4) To create context-sensitive AI implementation recommendations which help Zimbabwean banks and
regulators achieve their inclusion goals while reducing bias, exclusion and their decision-making processes
from being hidden.
Page 978
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
LITERATURE REVIEW
AI-Powered Credit Scoring and Access to Credit
The credit scoring system uses artificial intelligence to implement machine learning techniques which determine
creditworthiness through non-traditional data sources that include mobile phone usage and transaction records
and social behaviour patterns for people who do not have access to standard banking services (Nuka and
Ogunola, 2024). The research findings show that AI-assisted credit scoring systems improve risk evaluation by
decreasing information gaps which enables lenders to provide loans to people who lack conventional credit
records (Museba, 2024). However, issues of data privacy might become an area of great concern in such a
situation, leading to bias. In Zimbabwe, where people living in rural areas and low-income groups do not possess
official financial documents, AI-based credit systems have the capability to create new pathways for people to
access banking services (Kawara, Muchara, and van der Poll 2025).
Sub-Saharan Africa evidence shows that microloans and small business loans can have high approval rates using
an AI scorer, even for those borrowers with thin files, but with very low default rates (Thu, Doan, and Quy 2024).
However, it is necessary to analyse the capacity of the models. Museba (2024) reports that banks could use
alternative-data AI models to assess risk dynamically to increase credit access in new market segments initially
categorised as underserved. Kawara, Muchara, and van der Poll (2025) have emphasised that using AI within
Zimbabwean banks would very well transform the financial landscape for the rural poor since it boosts accession.
Nonetheless, these developments need to be tackled with care in order to improve inclusivity without widening
differences. The argument for AI-based scoring providing real-time decision-making, thereby enhancing the
speed and reliability of credit approvals, has been taken further by Thu, Doan, and Quy (2024). Additionally,
Museba (2024) suggests that lenders could be encouraged to give loans to excluded groups if AI predictive ability
reduces lender risk Therefore, we hypothesise that;
H₁: AI-powered credit scoring significantly improves access to credit for financially excluded groups in
Zimbabwe.
Chatbots, Virtual Assistants, and Usage of Financial Services
The introduction of AI-powered chatbots and virtual assistants into banking operations improved customer
engagement through personalised service delivery and faster service execution, which led to increased financial
service usage (Lee, 2024). The chatbot system enables customers to receive immediate support whilst learning
about banking services. This benefits people without financial literacy skills and branch access (Makore, 2024).
However, this can unintentionally strengthen available differences. According to Masuka (2025), recent research
from Zimbabwe demonstrates that customers now use virtual assistants for basic financial planning tasks and
enquiries, which leads to increased and more intense usage of financial services. Similarly, Nuka and Ogunola
(2024) support the view that virtual assistants simplify service delivery, which results in shorter service times
and higher customer satisfaction, thereby increasing the likelihood that customers adopt and continue using
financial products. Nevertheless, the emphasis on efficiency can lead to paying a blind eye to other important
issues such as financial education.
Museba (2024) argues that artificial intelligence chat systems build stronger customer trust in banking services,
which helps rural customers who have difficulty accessing banking services. Thu, Doan, and Quy (2024) found
that virtual assistants function as banking system guides for customers who need assistance with difficult banking
tasks. According to Makore (2024), organisations enable service access to all users through chatbot
implementation, which uses local languages and provides support for users with low literacy skills. The studies
prove that virtual assistants and chatbots create positive changes in financial behaviour for people in Zimbabwe
who use banking services. In light with these discussion, the researchers hypothesised that;
H2: Chatbots and virtual assistants significantly enhance the usage of financial services in Zimbabwe’s banking
sector.
Page 979
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Anomaly Detection Algorithms and Security in Financial Transactions
The most prominent application of AI in financial transaction security is the anomaly detection algorithm, which
can detect unusual patterns being a potential indication of fraud, money laundering, or unauthorised access
(Bajracharya, Khakurel, Harvey; Rawat, 2023). Recent evidence suggests that an anomaly detection system
through machine learning has enhanced the speed and accuracy of suspicious financial activities and reduced
financial risk at the institutional as well as personal levels (Mestiri, 2024). However, reliability and fairness of
the algorithms needs to be checked. Given that Zimbabwe is a country with increasing incidences of cybercrime
in digital banking associated with the increased uptake in mobile money transactions, anomaly detection would
offer much-needed protection for customer assets (Museba, 2024). It has been determined by Thu et al. and Doan
and Quy (2024) that AI algorithms trained on transaction patterns could detect anomalies in real time. This can
stop losses before they take place. However, this is dependent on the quality of training data in use.
Kawara, Muchara, and van der Poll (2025) found that automated anomaly detection boosts customer confidence
and aids regulatory requirements, leading to trust in banking systems. While this improvement is appreciated, it
is also important to analyse the extent of the trust. Automated anomaly detection influences the combination
with adaptive learning models that keep updating over time as fraud takes new forms (Makore, 2024). Integrating
anomaly detection into digital banking platforms relieves human security teams of the operational burden while
enabling monitoring at scale (Nuka and Ogunola, 2024). Museba (2024) provided evidence that banks which
use AI-based systems for fraud detection have successfully prevented financial losses from their transactions and
this demonstrates the security advantages of their operational methods. If one is to think otherwise, this study
has found it anomaly detection algorithms improve the protection of cash transactions. Therefore, the researchers
hypothesised that;
H3: Anomaly detection algorithms significantly enhance security in financial transactions.
Predictive Analytics, Robo-Advisors, and Financial Literacy
Predictive analytics and AI-based robo-advisors together create personalised financial solutions which enhance
clients' ability to understand their finances better (Lee 2024). However, even though this integration is highly
appreciated, there is still need for a deep analysis of the tools. The research shows that clients use predictive
analytics to foresee their upcoming financial situations while they acquire savings skills and debt management
abilities, which leads to improved financial choices (Bajracharya, Khakurel, Harvey; Rawat 2023). This skill
development is appreciated but there is risk of heavy reliance on the robo-advisors. Masuka (2025) explained
that robo-advisors with mobile and online banking connections provide users with an interactive learning
platform which enables them to discover personal finance, budgeting and investment strategies. Users who
receive predictive insights together with automated financial assistance show increased confidence in their
ability to handle financial products, which leads to improved financial literacy results, (Museba, 2024). The
researchers Kawara, Muchara, and van der Poll (2025) indicated that AI tools create personalised
recommendations by analysing individual behavioural patterns together with financial data. This gave insights
on the following hypothesis development;
H4: Predictive analytics and robo-advisors significantly enhance financial literacy in Zimbabwe.
METHODOLOGY
Research Philosophy and Approach
The research philosophy which was used is pragmatism. It emphasises on practical solutions and multiple
methods to deal with complex phenomena such as AI adoption in Zimbabwe's banking sector (Dube et al., 2024).
This philosophy supports the collection of both quantitative and qualitative data to investigate how AI affects
financial inclusion from various perspectives (Creswell & Plano Clark, 2023). Abductive research approach was
employed as it supports an iterative movement between observed data and theoretical constructs for the
generation of plausible explanations on how AI technologies, chatbots, anomaly detection, and predictive
Page 980
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
analytics-influence financial inclusion outcomes in a local banking context (Timmermans & Tavory, 2020;
Charles & McNulty, 2021).
Research Design
The present study adopted a mixed-method research design which used correlational and exploratory methods
to study how AI adoption relates to financial inclusion metrics while investigating the institutional and
operational processes that support those relationships (Creswell & Plano Clark 2023). The study established a
complete analysis which linked the generalisable results from quantitative research to the detailed contextual
information obtained from qualitative research (Bryman 2016; Field 2018).
Population, Sampling, and Sample Size
The total number of employees who worked in major commercial banks in Harare was 1754. This included 923
management personnel and 831 regular employees. The researchers used stratified random sampling to choose
their survey participants. Additionally, purposeful sampling was used to select the participants for the interviews.
The researchers used Raosoft to calculate a sample size of 293 respondents, which would provide enough
statistical power for quantitative research. Similarly 12 participants were used for qualitative data collection.
Data Collection Procedures
Researchers distributed the questionnaires through online platforms. They conducted semi-structured interviews
with twelve (12) management members both face to face and telephone interviews to explore their motivations,
the implementation challenges, and expected results of AI systems.
Data Analysis
The researchers conducted data cleaning procedures on the quantitative data set. The sample data and variable
characteristics were summarised using descriptive statistics. Quantitative data was analysed using Software
Package for Social Sciences (SPSS). Hypotheses testing was done using multiple regression analysis. The
researchers used NVivo software to conduct thematic analysis for qualitative data, which allowed them to
discover institutional practices and difficulties while studying how people view AI-based inclusive practices
(Braun and Clarke, 2021; Flick, 2022).
Validity and Reliability
The interview guide was developed from recent literature on AI and financial inclusion and received expert
reviews from banking and academic members. This established its content validity according to Cohen, Manion
and Morrison (2021) and Yegidis, Meyers and DePanfilis (2021).
The researchers established reliability for quantitative data through pilot testing and Cronbach alpha computation
which confirmed that all constructs maintained internal consistency above 0.7 according to (Taber 2018 and Hair
et al. 2020.
The research achieved qualitative trustworthiness through triangulation and member checking and thick
description which established the research findings' credibility, dependability and confirmability (Zamawe,
2015; Creswell & Plano Clark, 2023; Gill, 2021). The structured questionnaire was developed using validated
measurement scales in order to achieve content validity (Lincoln & Guba, 1985; Cohen, Manion & Morrison,
2021; Yegidis, Meyers & DePanfilis, 2021).
Ethical Considerations
The researchers protected participant confidentiality through total removal of personal identifiers. The study
participants received information about their right to choose whether to participate or to leave the study at any
moment. The researchers conducted the study with complete transparency which protected participants from
possible harm (Bryman 2016; Grix 2018; Saunders, Lewis & Thornhill 2019).
Page 981
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
RESULTS
Descriptive Results for AI in Banking
The table below indicated how respondents viewed AI technologies which were used by Zimbabwean banks for
credit scoring and fraud detection.
The study presented a complete assessment of AI technologies which include effectiveness, convenience,
security features, and obstacles they present to AI technology implementation. Table 1.1 illustrates the
descriptive statistics on AI in banking.
Table 1: Descriptive Statistics on AI in Banking
Statement
N
Mean
Std. Dev
AI improves loan approval chances
293
3.82
0.74
AI credit assessment is fair
293
3.68
0.79
AI credit decisions are transparent
293
3.25
0.84
AI expands access for unbanked
293
3.91
0.69
AI reduces bias in lending
293
3.54
0.81
I regularly use AI chatbots
293
3.45
0.83
Chatbots improve access
293
3.76
0.79
Virtual assistants increase convenience
293
3.89
0.71
Chatbots provide accurate responses
293
3.62
0.78
AI reduces branch visits
293
3.81
0.74
AI enhances online safety
293
3.91
0.71
AI helps detect suspicious activity
293
3.78
0.76
Predictive analytics aids decision-making
293
4.14
0.71
Robo-advisors improve financial understanding
293
3.89
0.74
AI recommendations improve budgeting
293
4.05
0.69
AI literacy tools should expand
293
4.29
0.64
Limited digital literacy is a barrier
293
4.21
0.66
Poor internet affects AI use
293
4.13
0.70
AI benefits urban more than rural users
293
4.10
0.75
High costs limit access
293
4.07
0.68
Source: Primary Data (2025)
The research results demonstrated that artificial intelligence technology delivers significant advantages for both
banking operations and financial accessibility efforts in Zimbabwe. The credit scoring and advisory tools
improve the access decision and financial literacy, which both need established solutions that function at lower
transparency (3.25) and trust in AI advice (3.62) standards as shown in Table 1 above.
Fraud detection and security instruments received high confidence ratings between 3.78 and 4.02, whilst chatbots
and virtual assistants provided users with greater convenience, which led to a decrease in the need for physical
bank visits between 3.45 and 3.89. The following barriers, limited digital literacy (4.21), internet connectivity
(4.13), costs (4.07), and urban-rural disparities (4.10), highlighted infrastructural and educational gaps which
required certain needs.
Page 982
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Hypothesis Testing
Normality Test (Shapiro-Wilk Test)
The Shapiro-Wilk test was used to determine whether the variables which measured AI adoption and its impact
on financial services operated under a normal distribution pattern. Table 2 below indicated the Shapiro-Wilk
Normality Test.
Table 2: Shapiro-Wilk Normality Test
Variable
Df
Sig.
AI credit scoring impacts credit access
293
0.001
Chatbots & virtual assistants improve service usage
293
0.004
Anomaly detection enhances transaction security
293
0.005
Predictive analytics & robo-advisors enhance financial literacy
293
0.003
Source: Primary Data, 2025
Table 2 above showed that all tested variables displayed normal distribution violations as shown by their p-value
results that fell below the 0.05 threshold. AI-enabled credit scoring showed the most significant deviation with
a p-value of 0.001 and chatbots or virtual assistants technologies showed a p-value of 0.004. The Central Limit
Theorem enables the study to use parametric methods since the researchers used a large sample size.
Correlation Analysis
The researchers used correlation analysis to measure both the strength and direction of relationships between AI
technologies and important financial results. Table 3 below illustrated the correlation matrix.
Table 3: Correlation Matrix
Variables
AI Credit
Scoring
Chatbots/Virtual
Assistants
Anomaly
Detection
Predictive Analytics/Robo-
Advisors
Access to Credit
0.695**
0.667**
0.648**
0.605**
Service Usage
0.672**
0.718**
0.621**
0.598**
Security
0.655**
0.639**
0.712**
0.623**
Financial Literacy
0.591**
0.578**
0.549**
0.732**
Note: p < 0.01 (2-tailed)
Source: Primary Data, 2025
Table 3 above highlighted that artificial intelligence systems produce positive outcomes which enhance the
performance of financial service providers. The most important relationship between AI credit scoring and credit
access showed a correlation of r = 0.695** while chatbots and virtual assistants show the highest connection to
customer service utilization with a correlation of r = 0.718**. Anomaly detection showed its strongest link to
security systems with a correlation of r = 0.712** whilst predictive analytics together with robo-advisors
demonstrated their strongest relationship to financial literacy with a correlation coefficient of r = 0.732**.
Page 983
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Multiple Regression Analysis
Multiple regression analysis was done to determine the unique contribution of each AI technology to improving
financial services. Table 4 below highlighted the multiple regression analysis results.
Table 4: Multiple Regression Analysis
Model
Unstandardized B
Standardized Beta
t
Sig.
(Constant)
1.134
-
4.215
0.000
AI Credit Scoring
0.265
0.298
5.492
0.000
Chatbots/Virtual Assistants
0.236
0.278
4.783
0.000
Anomaly Detection
0.301
0.342
6.123
0.000
Predictive Analytics/Robo-Advisors
0.247
0.272
4.725
0.000
Source: Primary Data, 2025
Table 4 above showed that all AI technologies delivered beneficial outcomes for financial service performance.
The security measure which protects transactions establishes anomaly detection as the most significant factor as
shown by the regression coefficients (Beta = 0.342, t-value = 6.123 and p<0.001).
AI credit scoring system had the following regression coefficients (Beta = 0.298, t-value = 5.462 and p< 0.001)
and it functioned as a primary AI technology which improved credit accessibility for users. Results indicated
that predictive analytics and roboadvisors were very significant as shown by their regression coefficients of (Beta
= 0.272, t-value = 4.725 and p< 0.001) since they assist users in acquiring financial knowledge to make superior
decisions.
The combination of chatbots and virtual assistants had (Beta = 0.278, t-value = 4.783, p<0.001) and enabled
users to obtain financial services. This gave insights on how AI adoption in Zimbabwean banks benefits customer
access, security, financial knowledge and customer satisfaction.
Regression Model Summary
The researchers used multiple regression analysis to evaluate how well AI technologies could explain differences
in key financial service results which included credit access, service usage, security measures and financial
literacy levels. Table 5 below showed the summary of the regression model.
Table 5: Regression Summary
Model
R
R Square
Adjusted R Square
Std. Error of Estimate
1
0.789
0.623
0.619
0.478
Source: Primary Data, 2025
The model demonstrated a high correlation between AI usage and financial service improvements which resulted
in an R value of 0.789 as shown in Table 5.
The R² value of 0.623 indicated that AI technologies explain 62.3% of the differences between financial service
outcomes while the adjusted R² value of 0.619 showed model stability after correcting the predictor variables.
The standard error value of 0.478 demonstrated that prediction results fall within the range of moderate accuracy.
Page 984
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Thematic Analysis
The qualitative results from this section of the study were obtained through semi-structured interviews with 12
senior managers at Zimbabwean commercial banks. The researchers used thematic analysis which resulted in
seven major themes being identified.
Theme 1: AI-Driven Credit Scoring and Financial Inclusion
“AI has enabled us to assess customers using alternative data, especially those without formal employment
records.” (Participant 1)
The response showed how AI technology decreases the need for standard credit evaluation methods which
depend on documents such as pay slips and collateral assets. The system demonstrated its capacity to provide
financial services to people who work in informal employment through its use of artificial intelligence.
“We are now comfortable lending to small traders because AI predicts repayment behaviour more accurately.”
(Participant 4)
This statement showed that financial institutions have developed more trust in their ability to lend money to
people who were once denied access to credit. The results showed that artificial intelligence technology
improved risk assessment results which organisations used to make lending decisions that promote financial
inclusion.
“Loan approvals are faster and more objective since human bias is removed.” (Participant 6)
The results demonstrated that AI systems increased operational productivity while preserving fair credit
distribution practices. The study highlighted that automated decision making systems established financial
systems which maintain transparent operations and provide equal access to all users.
Theme 2: Operational Efficiency through AI Automation
“Most of the companies rely on artificial intelligence for back-office work due to greater efficiency and economic
returns for the business in general while human staff gets closer to higher-value tasks of concentration and other
jobs.” (Participant 2)
This indicated that AI improves internal efficiency by reducing manual workload. Improved efficiency indirectly
enhanced service delivery to customers.
“One remarkable impact of AI on the processing speed of all activities is the reduction in the time it took to
complete a task.” (Participant 5)
This response showed how automation saves time for work performance. Faster processing times lead to better
customer satisfaction while increasing service access.
“Errors have been reduced because AI systems are consistent.” (Participant 8)
The results indicated that AI technology brings more accurate results and dependable performance to banking
workers. Digital financial services experience increased trustworthiness because of decreased errors.
Theme 3: AI Chatbots and Service Accessibility
“Indicates that Chatbots have facilitated banking services even during those times when they are normally off.”
(Participant 3)
Page 985
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
The system shows how artificial intelligence removes time-based limitations which prevent people from
accessing financial services. The system enables more customers to access its services who cannot travel to
physical locations.
“Customers now resolve issues instantly without visiting the bank.” (Participant 6)
The system shows better user convenience because it reduces the expenses of conducting transactions. The
system enables more people to start using digital banking services because it makes the services easier to access.
“Some customers are still adjusting, but chatbot usage is increasing.” (Participant 9)
The statement shows how people develop new behaviours through their experience with AI tools. The study
shows that people now accept the technology more than they did in the past.
Theme 4: Fraud Detection and Transaction Security
“AI detects suspicious transactions in real time, reducing fraud losses.” (Participant 1)
The study demonstrates how artificial intelligence improves transaction security through its effective security
enhancements. Digital platforms achieve higher customer trust through security improvements.
“Customers trust digital banking more because of AI monitoring.” (Participant 4)
The connection between AI security systems and improved trust. The trust relationship forms the basis which
enables people to access financial services.
“AI supports compliance with anti-money laundering regulations.” (Participant 7)
The answer describes how artificial intelligence implementation helps improve governance systems. The
implementation of strong compliance frameworks protects banking institutions and their customers through its
security advantages.
Theme 5: Predictive Analytics and Financial Literacy
“AI tools help customers plan their finances better.” (Participant 2)
Financial planning processes which use AI technology create a foundation for making correct financial decisions.
Organisations achieve continuous financial stability through enhanced planning processes which deliver better
financial results.
“Robotic assistance renders the abstruse data simple to understand.” (Participant 5)
The statement shows that AI reduces information asymmetry. The technology helps customers understand
financial matters better because it provides them with financial literacy training.
“The use of AI is supposed to unleash life and its dynamics from the ghetto of inherent sluggishness and
dullness.” (Participant 8)
The research shows that using artificial intelligence systems by themselves leads to specific operational
challenges. The study proves that financial education programs require development because standalone
artificial intelligence systems have operational restrictions.
Theme 6: Digital Literacy and Skills Constraints
“Digital illiteracy hinders customers from benefiting from AI.” (Participant 3)
The response identifies human capacity as the primary obstacle which prevents progress. The explanation
describes why different organisations adopt AI-enabled services at different rates.
Page 986
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
“Older customers struggle more with AI platforms.” (Participant 6)
The citation shows how various demographic groups adopt technology at different levels. The research shows
that vulnerable groups need specific support services which should be provided to them.
“Training customers is as important as deploying technology.” (Participant 9)
The statement emphasises that organisations need capacity building to develop their abilities. Users who can
effectively run AI systems acquire higher skills to work with all AI systems.
Theme 7: Infrastructure and Cost Barriers
“Poor internet connectivity affects AI service usage in rural areas.” (Participant 4)
The present condition shows that infrastructure disparities create obstacles which stop people from reaching
social inclusion. The rural population experiences more exclusion from society than other groups do.
“High costs prevent many people from using digital banking frequently.” (Participant 7)
The statement shows that people face difficulties in paying for essential things. The economic barriers which
exist at present prevent people from using artificial intelligence services which should be accessible to all users.
“Not everybody can afford smartphones required for an AI platforms.” (Participant 1)
Accessible assistive devices remain out of reach for users because of current distribution methods. The
development of inclusive AI requires implementation of socioeconomic interventions which extend beyond
current requirements for its development.
DISCUSSION OF RESULTS
The researchers found that their statistical findings confirmed the results of their descriptive research since they
discovered different statistical patterns between AI technologies and financial service outcomes. The normality
assumption for the variable gets violated when researchers use social science datasets which Ghasemi and
Zahediasl (2023) identified as having critical importance. The researchers demonstrated that large sample sizes
enabled studies to produce consistent results for parametrical tests which needed data from non-normally
distributed samples. The explanatory regression indicated that that AI predicts 62.3% of financial service
outcome variations which supports the theoretical expectation that AI acts as a major factor affecting modern
banking performance Pramanik et al. (2023). The correlation results showed that AI acts as a crucial element.
Credit scoring established the strongest connection to credit access, chatbots generated the most service usage
while anomaly detection helped handle security threats and, predictive analytics supported financial literacy
growth. Liu and Wang 2024, conducted a research which examined the multiple ways on how AI impacts
financial ecosystems. The regression analysis results showed that all AI elements functioned as strong predictors
for financial service enhancements, with anomaly detection showing particularly strong results concurred with
the results for Chang et al. 2022. This researcher propounded that security-based AI solutions generated the
highest economic value from their marginal advantages. The present study's hypotheses received strong support
from these findings.
The study investigated the relationship between AI-powered credit scoring system and access to credit for
financially excluded groups. The results showed a significant relationship between the two variables and this is
shown by the following regression coefficients (Beta = 0.298, t = 5.462, p < 0.001). These findings offers a
powerful contribution to the Resource-Based View (RBV) (Barney, 1991). The significant beta coefficient
indicates that AI credit scoring is a rare and valuable strategic resource rather than merely a tool. It develops a
new dynamic capability, the capacity to evaluate the creditworthiness of financially excluded populations by
utilizing non-traditional data. This transforms a traditional barrier into a new market opportunity, thereby
generating a competitive advantage that is difficult for competitors to imitate, fulfilling the VRIN criteria
Page 987
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
The researchers analysed the relationship between chatbots and virtual assistance and financial service usage in
Zimbabwe’s banking sector. Results from regression analysis indicated that there is a strong significant
relationship which is explained by a P-value = < 0.001, Beta = 0.278, t = 4.783, p < 0.001. This confirmation
supports Institutional Theory (DiMaggio & Powell, 1983) but moves it beyond mimetic isomorphism. Chatbots
are actually changing how banks operates, banks are not just following trends, they are setting new rules for how
customers interact with them, making 24/7 on-demand service the new normal. This proactively shapes the
market environment rather than just reacting to it.
The results highlighted that anomaly detection stands out as the strongest predictor in the model with a P-value
of 0.001, Beta = 0.342, t = 6.123, p < 0.001. This implied that there is a positive significant relationship between
the two variables. This supported the views of the Resource-Based View theory which valued security and
referred it as a strategic asset. Additionally, the results perfectly back up the Task-Technology Fit model
(Goodhue & Thompson, 1995). The technology matches the demands of securing financial transactions better
than anything else, and the strong regression numbers prove it.
The researchers tested fourth hypotheses on the relationship between predictive analysis, robo-advisors and
financial literacy in Zimbabwean banks. The results showed a highly positive relationship between the variables.
This relationship was supported by Beta = 0.272, t = 4.725, p< 0.001. This outcome intensely augments the
Institutional Theory. By meaningfully assisting users in acquiring financial knowledge, these technologies are
not just improving a service metric; they are acting as agents of normative isomorphism. They create and
disseminate new standards of financial understanding, thereby building a more financially literate population.
This positions banks as educators and institution-builders, leveraging AI to create long-term societal and
economic value, which in turn reinforces their own legitimacy and stability.
CONCLUSION
The research demonstrated that AI technologies have delivered major enhancements to Zimbabwean banks
through their capacity to improve credit access and customer service and transaction security and financial
literacy development. Study participants showed positive attitudes toward AI-based credit scoring, chatbots and
anomaly detection systems and predictive analytics because they considered AI technology to be a banking
service innovation that enhanced operational efficiency and accuracy and customer satisfaction. The public
requires better information about AI systems because their current knowledge prevents them from building trust
in these systems, which need transparent and consistent decision-making processes. The inferential analysis
results demonstrate that AI adoption functions as a strong predictor for financial service improvements because
the regression and correlation analysis results showed positive effects on all key outcomes. The study shows that
the AI credit scoring system successfully enhances credit access for borrowers while chatbot systems increase
service utilization and anomaly detection enables maximum protection for users. The primary purpose of
predictive analytics exists to enhance financial knowledge through its predictive capabilities. The research found
its initial hypotheses correct by proving that AI serves as an essential force which drives both financial inclusion
and digital banking success in the contemporary financial system. The research demonstrates that digital literacy
limitations, internet access problems, high-cost barriers, and urban–rural accessibility differences create multiple
structural and socio-technical obstacles which prevent the banking sector in Zimbabwe from achieving equal AI
implementation. The organisation needs to resolve these issues so that all customer groups can fully benefit from
AI development and participate in digital financial services which extends to customers from rural areas and
low-income brackets.
RECOMMENDATIONS
Recommendations for practice
Strengthen Transparency in AI Decision-Making: Banks must develop basic customer explanations about
their AI credit scoring systems which shows how data is used for credit decisions and offer customers ways to
improve their credit status. The solution solves current transparency issues which people have reported while
establishing trust in AI systems.
Page 988
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Expand AI-Driven Financial Literacy Tools: Financial institutions should increase the deployment of
predictive analytics and robo-advisory platforms targeted at low-income and first-time users. Financial literacy
tools need to be developed through basic applications which utilize local languages and mobile banking systems
to make them more accessible to users.
Improve Chatbot and Virtual Assistant Functionality: The existing banking chatbot systems need
enhancements because they do not provide customers with accurate answers to their inquiries throughout the
entire day. The implementation of voice-enabled features creates better accessibility for users who have reading
difficulties while matching the reading habits identified in the customer study.
Enhance AI Security Mechanisms: Anomaly detection systems create a major security risk for banks which
need to invest in fraud detection through real-time systems and biometric authentication and multi-layer
verification systems to protect against cyber threats. Digital transaction security needs to maintain high customer
confidence through regular security audits and customer alert systems.
Reduce Barriers to AI Use through Customer Training
The banking system requires financial institutions to develop digital literacy initiatives which should be
presented through online platforms and physical community centres to educate customers on using AI banking
systems. Digital skills development should be implemented in rural regions because it serves as the main obstacle
for people who want to access digital services.
Policy Recommendations (For Government and Regulators)
Develop a National AI Governance Framework for Banking: The Reserve Bank of Zimbabwe together with
other regulatory bodies must create specific rules which banks need to follow when adopting AI technology. The
regulations need to establish rules for four major areas which include algorithm transparency and data protection
and fairness and accountability to ensure proper ethical development of AI systems.
Invest in Digital Infrastructure for Rural Financial Inclusion: The government needs to focus on three main
areas which include enhancing internet access and developing better bandwidth systems and decreasing data
expenses for rural populations. The study shows that public-private partnerships can be used to build telecom
networks which brings services to remote areas.
Provide Incentives for Banks to Adopt Inclusive AI Technologies: The government should provide banks
with tax reductions and research funding and innovation testing facilities so banks can create artificial
intelligence systems that benefit people who experience financial difficulties particularly women and young
people and small business owners.
Strengthen Cybersecurity Regulations: Policymakers need to establish mandatory cybersecurity regulations
for AI-enabled banking systems which should include continuous auditing processes and breach reporting
systems and resilience testing procedures to combat digital fraud and build public trust in digital banking.
Promote Affordable Access Through Policy Interventions: The government must create policies which reduce
mobile data costs and provide free access to essential banking services because financial services need to be
available to all people who want to use AI technology for improved financial services.
Recommendations for Future Research
Research about AI usage needs to study how different age groups, gender, income brackets, rural and urban
residents adopt AI technology because it creates different effects which need specific solutions. The researchers
need to study whether AI-based credit scoring systems make borrowers better at repaying their loans while their
credit score improves and the system operates to decrease default rates through time. Additionally, future
research could incorporate perspectives from banking customers to complement the institutional insights
obtained from banking institutions.
Page 989
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Limitations of the Study
Researchers found that their study results didn't apply to all individuals in Zimbabwe because multiple limitations
restricted output generalisation. The research used a cross-sectional self-reported survey which introduced
response biases. The study included banking customers from selected urban and peri-urban areas which created
an incomplete picture of financial exclusion and digital access problems that exist in rural territories. The
research studied credit scoring chatbots, anomaly detection, and predictive analytics while excluding all other
available technologies that include biometric verification and block chain-based systems because those
technologies also affect financial inclusion. The banks implemented AI technology through various methods
which made it difficult for researchers to track all AI system changes because these systems continuously develop
new technological capabilities. The study's limitations force researchers to limit their findings to specific
financial institutions and demographic groups who exist in Zimbabwe.
Declaration of Conflict of Interest
The authors declared that there were no conflicts of interest associated with this study, and no external
organisation, institution, or funding body provided sponsorship or financial support for the research. The authors
conducted all data collection activities, analysis activities and reporting activities which they used to present
their study results while maintaining complete independent control over their research study.
REFERENCES
1. Alshamrani, A. (2024). AI-enhanced fraud detection: Novel approaches and performance evaluation. EAI
Endorsed Transactions on Security and Safety. https://doi.org/10.4108/eai.23-11-2023.2343170
2. Asongu, S. A., Biekpe, N., & Cassimon, D. (2024). Digital technologies and financial inclusion in
SubSaharan Africa. Telecommunications Policy. https://doi.org/10.1016/j.telpol.2022.102315
3. Bajracharya, A., Khakurel, U., Harvey, B., & Rawat, D. B. (2023). Recent advances in algorithmic biases
and fairness in financial services: A survey. EAI Endorsed Transactions on AI and Robotics.
https://doi.org/10.4108/airo.4252
4. Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1),
99–120. https://doi.org/10.1177/014920639101700108
5. Bracke, P., Datta, A., Jung, C., & Sen, S. (2021). Machine learning explainability in finance: An
application to default risk analysis. Bank of England Staff Working Paper No. 816.
6. Braun, V., & Clarke, V. (2021). Thematic analysis: A practical guide. SAGE Publications Ltd.
https://us.sagepub.com/en-us/nam/thematic-analysis/book248481
7. Bryman, A. (2016). Social research methods (5th Ed.). Oxford University Press.
https://global.oup.com/academic/product/social-research-methods-9780199689453
8. Chang, V., Golightly, L., Modesti, P., Xu, Q. A., Doan, L. M. T., Hall, K., Boddu, S., & Kobusinska, A.
(2022). A survey on the impact of artificial intelligence on organisational cyber security. Internet of
Things, 23, Article 100892. https://doi.org/10.1016/j.iot.2023.100892
9. Charles, V., & McNulty, Y. (2021). Premature return of self-initiated expatriates in the Middle East: A
study of motivations and barriers. The International Journal of Human Resource Management.
https://doi.org/10.1080/09585192.2023.2252317
10. Chinomona, E., & Sandada, M. (2022). Impact of digital finance on financial inclusion in Zimbabwe:
Africa's quest for inclusion trends and patterns. African Journal of Economic and Management Studies.
https://www.researchgate.net/publication/385417699_Impact_of_digital_finance_on_financial_inclusio
n _in_Zimbabwe_Africa's_Quest_for_Inclusion_Trends_and_Patterns
11. Cohen, L., Manion, L., & Morrison, K. (2021). Research methods in education (8th Ed.). Routledge.
https://www.routledge.com/Research-Methods-in-Education/Cohen-
ManionMorrison/p/book/9781138209886
12. Creswell, J. W., & Plano Clark, V. L. (2023). Designing and conducting mixed methods research (3rd
Ed.). SAGE Publications.
13. Dube, B., Nkomo, D., & Apadile-Thokweng, M. (2024). Pragmatism: An essential philosophy for mixed
methods research in education. International Journal of Research and Innovation in Social Science, 8(3),
1001-1010. https://doi.org/10.47772/IJRISS.2024.803076
Page 990
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
14. DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and
collective rationality in organizational fields. American Sociological Review, 48(2), 147–160.
https://doi.org/10.2307/2095101
15. Field, A. (2018). Discovering statistics using IBM SPSS statistics (5th Ed.). SAGE Publications.
16. Flick, U. (2022). An introduction to qualitative research (7th Ed.). SAGE Publications Ltd.
https://uk.sagepub.com/en-gb/eur/an-introduction-to-qualitative-research/book278983
17. Ghasemi, A., & Zahediasl, S. (2023). Normality tests for statistical analysis: A guide for nonstatisticians.
International Journal of Endocrinology and Metabolism, 10(2), 486-489.
https://doi.org/10.5812/ijem.3505
18. Gill, S. L. (2021). Qualitative research methodology: Application of triangulation. Nurse Researcher,
29(2), 5-6. https://doi.org/10.7748/nr.29.2.5.s1
19. Goodhue, D. L., & Thompson, R. L. (1995). Task-technology fit and individual performance. MIS
Quarterly, 19(2), 213–236. https://doi.org/10.2307/249689
20. Grix, J. (2018). The foundations of research (3rd Ed.). Bloomsbury Publishing.
https://www.bloomsbury.com/us/foundations-of-research-9781352002003/
21. Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2020). Multivariate data analysis (8th Ed.).
Cengage Learning. https://www.cengage.com/c/multivariate-data-analysis-8e-hair/9781473756540
22. Hossain, M. A., Yesmin, M. R., Rahman, M. S., & Jahan, N. (2023). Advanced fraud detection using
machine learning models: Enhancing financial transaction security. Journal of Computer Science and
Technology Studies, 5(2), 1-12. https://doi.org/10.32996/jcsts.2023.5.2.1
23. Kawara, A., Muchara, B., & van der Poll, H. M. (2025). Financial inclusion in Zimbabwe: Lessons from
the commercial banking sector. International Journal of Financial Studies, 13(4), Article 182.
https://doi.org/10.3390/ijfs13040182
24. Khan, M. K., Trinh, H. H., Khan, I. U., & Ullah, S. (2023). The adoption of conversational assistants in
the banking industry. Heliyon, 9(9), e17428. https://doi.org/10.1016/j.heliyon.2023.e17428
25. Kshetri, N. (2023). AI-powered credit scoring models: Ethical considerations, bias reduction, and
financial inclusion strategies. International Journal of Research Publication and Reviews, 6(3), 1-10.
https://ijrpr.com/uploads/V6ISSUE3/IJRPR40581.pdf
26. Lee, J. (2024). AI-driven approaches for real-time fraud detection in US financial transactions:
Challenges and opportunities. ResearchGate.
https://www.researchgate.net/publication/381548442_AIDriven_Approaches_for_Real-
Time_Fraud_Detection_in_US_Financial_Transactions_Challenges_and_Opportunities
27. Li, J., & Xu, J. (2022). Deep learning for financial forecasting: A review of recent trends. International
Review of Financial Analysis. https://doi.org/10.1016/j.irfa.2024.103490
28. Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. SAGE Publications.
https://us.sagepub.com/en-us/nam/naturalistic-inquiry/book842
29. Liu, Y., & Wang, W. (2024). The value-creating potential of AI: A multi-dimensional analysis of effects
and mechanisms. International Review of Financial Analysis, 91, Article 103057.
https://doi.org/10.1016/j.irfa.2024.103057
30. Makore, M. (2024). Regulating artificial intelligence to advance financial inclusion in South Africa.
Potchefstroom Electronic Law Journal, 27, 1-31. https://doi.org/10.17159/1727-3781/2024/v27i0a17488
31. Maseeh, J., & Devi, A. (2022). The impact of AI-based chatbots and virtual assistants on customer
experience in digital banking. ResearchGate.
https://www.researchgate.net/publication/396689606_The_Impact_of_AI-
Based_Chatbots_and_Virtual_Assistants_on_Customer_Experience_in_Digital_Banking
32. Masuka, B. (2025). Determinants of financial inclusion among smallholder farmers in Zimbabwe.
Journal of Agricultural and Rural Development, 1(1), 1-15. https://doi.org/10.17306/j.jard.2025.00016r2
33. Mestiri, M. (2024). Detecting anomalies in financial data using machine learning algorithms. Systems,
10(5), Article 130. https://doi.org/10.3390/systems10050130
34. Mhlanga, D. (2023). Factors driving artificial intelligence adoption in South Africa's financial services
landscape. Academic Journal of Interdisciplinary Studies, 13(5), 1-15. https://doi.org/10.36941/ajis2024-
0135
35. Museba, T. (2024). Leveraging artificial intelligence for enhanced credit risk. In Advances in Intelligent
Systems and Computing (pp. 201-215). Springer. https://doi.org/10.1007/978-3-031-74050-3_15
Page 991
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
36. Nuka, P. E., & Ogunola, A. A. (2024). AI and machine learning as tools for financial inclusion:
Challenges and opportunities in credit scoring. International Journal of Scientific Research and
Engineering Development, 7(5), 1-12. https://ijsra.net/sites/default/files/IJSRA-2024-2258.pdf
37. Pramanik, S., Bandyopadhyay, S. K., & Ghosh, S. (2023). Transforming banking: Examining the role of
AI technology innovation on financial performance and risk management. International Review of
Financial Analysis. https://doi.org/10.1016/j.irfa.2024.103632
38. Saunders, M. N. K., Lewis, P., & Thornhill, A. (2019). Research methods for business students (8th Ed.).
Pearson. https://www.pearson.com/uk/educational/research-methods-for-business-students-8e.html
39. Sironi, P. (2023). Financial literacy, robo-advising, and the demand for human financial advice: Evidence
from Italy. Behavioural Public Policy. https://doi.org/10.1017/bpp.2023.23
40. Taber, K. S. (2018). The use of Cronbach's alpha when developing and reporting research instruments in
science education. Research in Science Education, 48(6), 1273-1296. https://doi.org/10.1007/s11165016-
9602-2
41. Thu, T. N., Doan, Q. C., & Quy, N. L. D. (2024). An experimental study on fairness-aware machine
learning for credit scoring problem. arXiv preprint arXiv:2412.20298.
https://doi.org/10.48550/arXiv.2412.20298
42. Timmermans, S., & Tavory, I. (2020). Data analysis in qualitative research: Theorizing with abductive
analysis. University of Chicago Press.
43. Tsan, M., Lee, B. W., & Hall, R. (2023). The effects of artificial intelligence applications in educational
settings. International Review of Financial Analysis. https://doi.org/10.1016/j.irfa.2023.102761
44. Yegidis, B. L., Meyers, R. S., & DePanfilis, D. (2021). Research methods for social workers (9th Ed.).
Pearson. https://www.pearson.com/en-us/subject-catalog/p/research-methods-for-
socialworkers/P200000002052/9780136912286
45. Zamawe, F. C. (2015). The implication of using NVivo software in qualitative data analysis:
Evidencebased reflections. Malawi Medical Journal, 27(1), 13-15. https://doi.org/10.4314/mmj.v27i1.4