Reimagining

Credit with AI
Executive
Summary

Key insights and takeaways

AI-driven credit scoring is reshaping the global financial landscape, enabling access
to credit for underserved populations, including the unbanked and underbanked. By
leveraging machine learning (ML), natural language processing (NLP), and
alternative datasets, financial institutions are democratizing credit, enhancing
inclusion, and improving risk assessment. This report highlights global case studies
and key lessons, focusing on AI's transformative potential in financial inclusion.

Scope and objective of the report

This report explores the global impact of AI in credit scoring, examining successful
implementations by leading financial institutions. It identifies lessons learned and
provides actionable recommendations to drive scalability, inclusivity, and
operational efficiency.

Reimagining Credit with AI 02

1. Introduction to
credit scoring
1.1 Evolution of credit scoring systems

Globally, traditional credit scoring models like FICO (Fair Isaac Corporation) have struggled
to cater to informal economies, where the lack of formal financial records excludes large
populations. Emerging markets, home to a significant portion of the unbanked, highlight
these limitations.
Static data dependency: Credit card transactions, bank account histories, and loan
repayments are insufficient in cash-based economies.
Exclusion of informal sectors: Gig workers, small business owners, and rural
populations are often left out due to rigid requirements.

FICO SCORE

10%
New Credit
30%
Amounts Owed
15%
Length of
Credit History

10%
Credit Mix

35%
Payement History
Source: myFICO

Reimagining Credit with AI 03

In SEA, where 70% of the population remains either unbanked or underbanked (World Bank,
2023), these data points are scarce. Many individuals and small businesses operate in cash-
dominated environments, leaving little to no digital or formal financial footprint.

Today

>70%
Consumers
Underbanked or unbanked today

88% 67%
of industry experts of industry experts
believe digital FS will believe unbanked will
improve access for the still have limited
underbanked by 2025 access to FS by 2025

1.2 Challenges in traditional credit scoring models

Data gaps: Over 1.4 billion adults globally lack access to formal financial services (World
Bank, 2023).

Source: World Bank

Reimagining Credit with AI 04

 Exclusionary practices: Rigid requirements often exclude gig workers with inconsistent
income patterns, small business owners, and rural populations.
Lack of flexibility: Credit scoring models like FICO, developed in markets like the U.S.,
often fail to incorporate cultural and behavioral factors relevant to SEA. For example, in
rural Indonesia, over 50% of households rely on informal credit sources, making
traditional systems irrelevant for assessing their creditworthiness.

More than 70% of Southeast Asia’s population is either underbanked or unbanked

Unbanked: Southeast Asia adult population 2018 (million)

No access to basic financial
400 5 23 55 181 67 70
services (a bank account)
2%
15%
18%
Underbanked:
Not well-served in financial 198
38% 51%
services or have unmet needs (50%)
65%
40% 69%
No acccess to credit cards, 45%

underinsured, no long-term
savings products 98
(24%) 26%
13%
Banked: 60% 10%
45%
Well-served in financial services 104 37%
(26%)
needs 23% 22% 21%

Have access to credit cards,

insured or have investment SEA Singapore Malaysia Thailand Indonesia Phiilippines Vietnam

products
Source: economysea

While global digital ecosystems are rapidly advancing, traditional scoring models remain
reliant on static data sources. In contrast, alternative data points such as mobile phone
usage, utility payments, and e-commerce activity are:

More accessible in SEA due to widespread mobile penetration (90% mobile penetration
in Indonesia, for example).
Strong indicators of creditworthiness for those excluded from traditional banking.

Traditional systems like FICO lack the capability to ingest and process such diverse data
types at scale.

Reimagining Credit with AI 05

2. The role of AI
in credit scoring
Artificial intelligence (AI) has emerged as a game-changing tool in credit scoring, particularly
in regions like Southeast Asia (SEA), where traditional systems often fall short. By leveraging
advanced technologies such as Machine Learning (ML), Natural Language Processing (NLP),
and Predictive Analytics, AI enables financial institutions to develop dynamic, inclusive, and
accurate credit scoring models.

Digital lending continues to grow steadily as an appealing and accessible option, particularly
for underserved communities with limited access to traditional financial services. Peer-to-
peer lending platforms have improved credit management practices, keeping non-
performing loan (NPL) ratios low.

2.1 Overview of AI technologies in credit analysis

Machine Learning (ML): Identifies hidden patterns in diverse datasets to predict

creditworthiness. Unlike traditional systems that rely solely on static credit reports, ML
incorporates alternative data like mobile phone usage, e-commerce activity, and utility
payments to assess creditworthiness.

Natural Language Processing (NLP): NLP processes and interprets unstructured data
such as customer service interactions, loan application narratives, or even social media
posts to generate additional insights into a borrower’s behavior and intentions.

Predictive Analytics: Predictive analytics uses historical data and AI algorithms to

forecast a borrower’s future financial behaviour. It evaluates variables such as income
trends, spending habits, and repayment timelines to determine risk levels.

Reimagining Credit with AI 06

 Machine Learning

AI Technologies
in Credit
Analysis

Natural Language Predictive

Processing Analytics

2.2 Benefits of AI in modern credit scoring

Broader inclusion: Alternative data allows scoring of individuals without traditional

financial histories.
Improved accuracy: AI models minimize biases and predict risks with greater
precision.
Dynamic decision-making: Enables real-time adjustments based on user behavior
and economic shifts.

The integration of AI in credit scoring is critical for a regions, where informal economies
dominate, and access to credit is limited. With alternative data, AI bridges the gap
between financial institutions and underserved populations, fostering financial inclusion
and driving regional economic growth. AI technologies, tailored to local challenges, are
proving instrumental in reshaping the global credit landscape.

Reimagining Credit with AI 07

3. Banks and fintechs
expand lending with AI
Key players in the digital economy have entered the DFS (Digital Financial Services) sector
by launching digital banks. They are embracing generative AI to power virtual assistants and
perform credit scoring for underserved customers, among other uses. These innovations are
set to boost speed and accuracy, intensifying competition in the financial landscape.

Case studies of successful implementations of AI in

credit scoring by financial institutions

CASE STUDY 1

Bank BRI, Indonesia – Microfinance with AI

In Indonesia, one of the biggest barriers to financial inclusion is the lack of formal credit
histories for many individuals, particularly those in rural areas. Bank BRI (Bank Rakyat
Indonesia) wanted to extend financial services, specifically microloans, to rural
populations who had no access to traditional credit systems. The issue was how to
accurately assess the credit risk of these rural borrowers who lacked a formal financial
footprint.

Challenge: Scoring rural borrowers without formal credit histories and to to facilitate
consumer credit digitally, without the need to visit branches.
Solution: AI-driven credit scoring using agricultural data such as crop yield patterns
and market prices with the help of digital lending apps like Pinang and Ceria. For
example, crop yield patterns could be linked with income projections based on historical
performance, while fluctuations in market prices for certain agricultural products helped
model financial stability.
Outcome: The AI-driven credit scoring system enabled Bank BRI to provide microloans
to cover million rural borrowers, many of whom had never before had access to formal
credit. Moreover, by using agricultural data and AI-powered risk assessment, the bank
managed to reduce default rates.

Reimagining Credit with AI 08

This is illustrated by the over 50,000 loan applications worth more than USD 8 Million
processed by the two apps. The loan volume across the two apps has also increased
significantly as Pinang has seen 62% increase while Ceria has disbursed IDR 4 billion of
loans every week, increasing by 45% per week. Another benefit to the bank is the reduction
in turnaround time enabling an increase in business.

Lessons learned:
Collaborations with local data providers, such as cooperatives, can yield valuable
alternative datasets for credit scoring. By leveraging local data instead of relying
solely on credit reports or bank statements, Bank BRI accessed previously
unavailable data that better reflects borrowers' actual economic activity and
income potential.
Custom AI models tailored to specific industries, like agriculture, enhance
precision and provide more accurate risk assessments for underserved groups.

o Achieved a reduction in the time taken to process loan applications via mobile app
from 2 weeks to less than 2 days.

o Enhanced fraud detection, reducing the rate of fraud by 40% to record low levels,
versus other banks.

A key lesson from this

case study is the
importance of integrating
external factors like
weather patterns,
market prices, and
environmental risks into
credit scoring models.

Source: Unsplash

Reimagining Credit with AI 09

CASE STUDY 2

Grab Financial, Southeast Asia – Gig economy credit scoring

The gig economy in Southeast Asia (SEA) has seen rapid growth, with millions of
individuals working as drivers, food delivery partners, and other freelance roles through
platforms like Grab. However, one of the major hurdles for these workers in accessing
financial services is the inconsistent nature of their income streams.

For Grab Financial, a subsidiary of Grab, this presented a significant challenge. The
company wanted to offer financial products like personal loans, microloans, and
insurance to its growing base of gig economy workers.

Challenge: Providing loans to gig workers with inconsistent income streams.

Solution: ·Leveraged platform data, including ride volume, customer ratings, and
transaction histories, to assess creditworthiness.

Ride volume: The number of rides or deliveries a driver completes, providing

insight into their activity level and potential earnings.
Customer ratings: The ratings given by passengers and clients after each ride
or delivery, which reflect the quality of service and customer satisfaction.
Transaction histories: Data from GrabPay and other in-app payment systems,
providing insights into the financial behavior and reliability of workers.

Outcome

High adoption rate: One in three active driver-partners has taken out a loan
through Grab, indicating the usefulness and accessibility of these financial
products.
Financial inclusion: For many driver-partners, taking a loan from Grab is their
first loan from a formal institution, helping to build their financial history.

The outcome was a win-win: Grab

Financial expanded its user base and
increased customer loyalty by offering
accessible financial products, while
gig workers gained access to loans
that allowed them to manage personal
expenses, invest in their businesses,
or improve their livelihoods.

Source: Grab

Reimagining Credit with AI 10

 Lessons learned:
Internal platform data, such as user activity and performance metrics, can serve as
reliable credit indicators for non-traditional workers like gig workers.
Continuous refinement of AI models is crucial to account for fluctuations in gig
economy behaviors and maintain low default rates.

With a variety of loan products tailor-made for gig-workers, Grab loan disbursals
increased 57% year-on-year to USD 1.5 billion in 2023.

The key lesson from Grab’s success is that internal platform data, such as ride volume,
customer ratings, and transaction histories, can serve as reliable indicators of creditworthiness
for non-traditional workers like gig workers. These data points are more representative of a
worker's income-generating potential than traditional credit scoring factors.

CASE STUDY 3

Tonik Bank, Philippines – Digital-first AI models

The Philippines is one of the fastest-growing digital economies in Southeast Asia,
characterized by a mobile-first population where smartphones are the primary means of
accessing the internet. Despite this digital adoption, a large portion of the population
remains unbanked or underbanked, with limited access to traditional financial services.

Tonik Bank, a neobank, sought to address this gap by providing accessible financial
products, such as personal loans and savings accounts, to underserved segments.
However, the lack of traditional credit histories and reliance on cash-based transactions
posed a significant challenge. Scaling a credit scoring solution that could accurately assess
the creditworthiness of these individuals was critical to expanding financial inclusion.

Challenge: Scaling credit scoring for a mobile-first population with limited access
to traditional banking.
Solution: AI-based dynamic risk analysis integrating e-commerce and telecom data for
personalized credit scoring.
E-Commerce data: Purchasing patterns, transaction volumes, and payment
histories from online shopping platforms. Regularity of payments and cart
abandonment rates to gauge financial behavior and reliability.
Telecom data: Mobile phone usage, such as call and text frequency, prepaid balance
top-ups, and data consumption trends. Bill payment histories for postpaid users,
providing insight into financial discipline and stability.

Reimagining Credit with AI 11

Tonik also partnered with
telecom providers to access
scalable data pipelines,
ensuring that even individuals
in rural or semi-urban areas
could be assessed using
reliable data. This mobile-
first approach aligned with
the digital habits of the
population, making the
process seamless and
accessible for users.

Source:prnasia

Tonik Bank collaborated with FinScore, a leading credit scoring company in the
Philippines. Since its launch, FinScore's technology has powered credit assessments for
over 3.5 million Filipinos and facilitated the disbursement of more than $500 million in
loans.

Outcome:
Increased financial inclusion: By leveraging FinScore's telco data credit
scoring, Tonik has been able to extend credit to previously underserved
populations, including those without traditional credit histories.
Tonik has built a $20 million USD lending portfolio, demonstrating the
effectiveness of FinScore's credit scoring technology.
Lessons learned:
Mobile-first strategies are highly effective for reaching digitally native
populations, particularly in emerging markets.
Collaborations with telecom companies can provide a scalable and accessible
data pipeline for real-time credit scoring.

Tonik Bank’s innovative use of AI-driven credit scoring showcases how alternative data
and mobile-first strategies can transform financial inclusion in emerging markets like the
Philippines. By integrating e-commerce and telecom data, Tonik was able to overcome
the limitations of traditional credit scoring systems, enabling access to credit for
underserved populations.

Reimagining Credit with AI 12

CASE STUDY 4

Kredivo, Indonesia – AI-driven credit for digital purchases

In Indonesia, e-commerce has become one of the most dynamic sectors, with millions of
consumers shifting to online platforms for their purchasing needs. However, a significant
portion of these consumers lacks access to formal credit systems due to the absence of
traditional credit histories. This gap makes it challenging for e-commerce platforms and
financial institutions to offer credit-based services, such as buy now, pay later (BNPL) or
personal loans, without incurring high risks.

For Kredivo, a leading fintech company in Indonesia, the primary challenge was to create a
system that could assess the creditworthiness of these consumers quickly and accurately.

Challenge: Offering credit to consumers with little or no traditional credit history,

particularly in the e-commerce space.

Solution: Implemented an AI-based platform using real-time data, including online

shopping behavior, payment histories, and mobile phone activity, to assess
creditworthiness.

Outcome: Approved loans for over 2 million users, with a 20% month-on-month
increase in approvals for individuals with no formal credit history.

Lessons Learned:
E-commerce platforms can serve as rich data sources for credit scoring,
reflecting a customer's reliability based on their purchasing patterns and
payment behavior.
AI algorithms must be tailored to assess unique consumer behaviors in fast-
paced, digitally driven sectors like online shopping.

By leveraging e-commerce behavior,

payment histories, and mobile
activity, Kredivo successfully bridged
the gap between digital consumers
and formal credit systems, enabling
millions of underserved users to
access credit safely and responsibly.

Source: asiatechdaily

Reimagining Credit with AI 13

 Kredivo's game-changing algorithm approves loans of up to 30 million rupiah
($2,100) in just two minutes, bypassing traditional credit scores in favor of
innovative data insights like smartphone type. With 90% of Indonesian customers
paying on time—matching traditional credit card repayment rates—Kredivo proves
that smarter, faster lending is not only possible but highly effective.

This case study underscores the importance of tailoring AI models to specific industries,
such as e-commerce, to maximize their effectiveness. For other fintechs looking to enter
similar markets, Kredivo’s approach offers a blueprint for leveraging data partnerships,
dynamic risk assessment, and real-time decision-making to create inclusive, scalable, and
sustainable credit solutions.

CASE STUDY 5

Timo Bank, Vietnam – AI-powered credit scoring for

freelancers and gig workers
Vietnam’s gig economy has grown exponentially, with a large number of workers in
freelancing, ride-hailing, delivery services, and other flexible employment sectors.

For Timo Bank, Vietnam's first digital bank, this presented a critical challenge:
1. How to assess the creditworthiness of gig workers and freelancers whose income
varies month-to-month.
2. How to offer customized financial solutions that align with the financial realities of
this growing segment.

Challenge: Scoring freelancers and gig workers, who typically have irregular
income and no formal credit history.
Solution: Timo implemented AI algorithms that utilized data from social media activity,
online purchases, and mobile transaction history to predict repayment behavior.

The AI models incorporated machine learning (ML) algorithms that dynamically adjusted to
reflect changes in gig workers' income and behavior. For instance:
Seasonal variations in income were factored into risk assessment for freelancers
with irregular projects.
Borrowers with increasing activity on professional platforms were flagged for
potential credit line expansions.

Reimagining Credit with AI 14

 Outcome: AI model will help Timo scale its business and acquire 5 million new
customers over the next three years. Timo offers personalized financial services to
customers meeting their individual financial needs in a secure and compliant manner.

Lessons learned:
Data points beyond traditional financial records, such as social media activity or e-
commerce purchases, can offer valuable insights into a borrower’s reliability.
AI models must be flexible enough to account for the fluctuating income streams
common to gig workers, allowing for dynamic loan terms that fit their financial
cycles.

By leveraging social media

activity, e-commerce behavior,
and mobile transactions, Timo
created a credit assessment
system that not only expanded
access to credit but also
maintained low default rates.

Source: Timo

“Our ultimate mission is to enable financial inclusion for everyone – not just for those
who are ‘unhappily banked’ or those who are currently ‘unbanked’, but inclusion across
our entire society. Utilizing Mambu’s cloud banking platform allows us to build a bank
that is truly accessible to everybody. We want to become a financial partner to our
customers and provide services that enhance our customers’ lives, and with Mambu
we’ve found a technology partner that allows us to do that.”

Henry Nguyen, CEO, Timo.

With its innovative approach, Timo Bank has set a benchmark for how digital-first financial
institutions can empower underserved segments while fostering financial inclusion in the gig
economy.

Reimagining Credit with AI 15

CASE STUDY 6

DBS Bank, Singapore – AI industrialization in credit scoring

In Singapore, DBS Bank sought to leverage AI across its banking operations to address
challenges in credit scoring and enhance customer experiences. As a leading financial
institution in Asia, DBS needed to ensure its AI-driven initiatives were scalable and
impactful across diverse banking functions.

Challenge: Integrating AI into credit scoring to enhance accuracy and decision-making,

while also improving operational efficiency and customer engagement.

Solution: DBS deployed over 800 AI models across 350 use cases, encompassing areas
such as credit risk assessments, customer personalization, and predictive analytics.
For credit scoring specifically, AI models were used to analyze an extensive range of
data points, including:

Transactional histories: Analysis of account inflows and outflows.

Behavioral patterns: Insights into spending habits, repayment behaviors, and
lifestyle indicators.
Alternative data sources: Information from digital footprints, social data (where
compliant with regulations), and economic indicators.

Generative AI tools are being explored to automate and refine critical processes like
credit approval memos and risk analysis.

Outcome: The AI implementation resulted in reduced non-performing loans (NPLs),

improved customer engagement, and significant operational efficiencies. DBS
anticipates over SGD 1 billion in economic impact by 2027, driven by AI-led cost
savings and increased revenue streams.

The bank developed over 100 algorithms analyzing 15,000 customer data points for
personalized financial advisory. DBS's AI initiatives generated SG$370 million in
economic value in 2023, more than doubling from the previous year.

Reimagining Credit with AI 16

 Lessons learned:
Diversifying data sources, including transactional and behavioral data, is essential
for robust credit scoring models.
Generative AI offers potential in automating complex decision-making processes,
such as credit approvals.
Scaling AI across operations requires a clear strategy for model deployment and
integration into existing workflows.

By industrializing AI, DBS Bank

has positioned itself as a
leader in innovation,
demonstrating how financial
institutions can leverage AI to
transform risk management
and customer engagement
while achieving substantial
economic gains. Other banks
can adopt DBS's approach to
harness AI for comprehensive,
scalable, and sustainable
digital transformation.

Source: Bloomberg

Reimagining Credit with AI 17

CASE STUDY 7

MYBank, China – AI-powered inclusion for SMEs

In China, small and medium enterprises (SMEs) face significant challenges in accessing
traditional credit due to limited credit histories, inadequate documentation, and high risk
perceptions from lenders. MYBank, an affiliate of Ant Group, recognized this gap and
sought to address it using artificial intelligence (AI). Their innovative approach
revolutionized SME lending, making credit more accessible and inclusive.

Challenge: In China, SMEs play a crucial economic role yet struggle to access credit due
to insufficient credit histories, high-risk perceptions, and resource-intensive traditional
evaluation methods.

Solution: MYBank harnessed AI-driven innovations to transform SME lending:

AI-powered credit scoring: Leveraged transactional data, such as sales

volumes and payment behaviors.
Integrated supply chain data, including vendor relationships and payment
cycles.
Used alternative data points, such as inventory levels and business activity on
e-commerce platforms, to create a more comprehensive credit profile.

Real-time loan approvals:

AI algorithms process millions of loan applications in real-time.
Loan approvals occur in seconds, with disbursements completed in under 3 minutes,
offering SMEs instant access to working capital.

MYBank utilizes the "Zhima credit evaluation system," an automated, AI-powered credit
scoring system that analyzes over 3,000 variables to assess an applicant's ability to repay
within three minutes.

Source: qorusglobal

Reimagining Credit with AI 18

 MYBank pioneered the "310 model" for SME financing: loans take less than three
minutes to apply on a mobile phone, less than one second to approve, and require
zero human interaction.

Scalable technology infrastructure:

Built a robust digital lending platform that could support high transaction volumes
without compromising accuracy or speed.
Emphasized accessibility through mobile platforms, ensuring inclusivity for SMEs in
remote areas.

By the end of 2023, MYBank had served over 53 million SMEs cumulatively,
demonstrating the scale of its impact on financial inclusion. Despite serving high-risk
segments, MYBank maintains a default rate of just 1%, showcasing the effectiveness of
its AI-driven credit assessment.

CASE STUDY 8

UnionBank, Philippines – AI for gender-inclusive lending

UnionBank of the Philippines has been at the forefront of using technology to drive
innovation in financial services. Recognizing the challenges faced by women entrepreneurs
in accessing credit, the bank leveraged artificial intelligence (AI) to create a more inclusive
lending ecosystem.

Challenge: Women entrepreneurs in the Philippines play a vital role in driving

economic growth but encounter systemic barriers to accessing credit, such as
biases in traditional scoring methods that undervalue their businesses,
underserved financial needs due to a lack of tailored products, and the absence of
alternative data in credit assessments. UnionBank's challenge was to address
these issues by designing a credit system that promotes inclusivity and fairness
while maintaining efficiency and scalability.

Reimagining Credit with AI 19

Source: connectedwomen

Solution: UnionBank deployed AI-powered solutions to design gender-sensitive credit

scoring algorithms aimed at empowering women entrepreneurs:
Alternative data integration: AI models analyzed non-traditional data sources, such
as:
Business performance metrics: Revenue growth, profit margins, and cash flow
stability.
Spending patterns: Trends in business-related and personal expenditures.
Behavioral data: Transactional behaviors and repayment histories specific to
women-led businesses.
Outcome: A significant rise in credit approvals for women-led businesses, enabling
them to access the capital needed for growth and expansion. UnionBank’s efforts
contributed to bridging the gender gap in access to formal financial services in the
Philippines.

UnionBank's fintech arm, UBX, partnered with Singapore-based Bixie Pte. Ltd. to create
an open finance platform targeting women. This partnership aims to disburse aid funds
directly to Filipino women beneficiaries through a "last mile" disbursement network.

Solution:
Tailoring AI algorithms to address biases and reflect the unique needs of
underserved groups is crucial for equitable credit systems.
Understanding gender-specific financial behaviors and needs drives the success of
inclusive lending initiatives.

Reimagining Credit with AI 20

4. Five lessons from
emerging markets in SEA
Mobile-first strategies cater effectively to digitally
01
native populations:
Emerging markets like SEA have high smartphone penetration but low banking access.
Mobile-first credit solutions meet users where they are, offering seamless and inclusive
financial services.

Partnerships with telecom companies provide a

02
scalable data pipeline for credit scoring:
Telecom companies maintain extensive datasets (e.g., call patterns, bill payments) that
serve as reliable indicators for creditworthiness, enabling financial institutions to scale
credit scoring efficiently.

Leveraging non-traditional datasets in credit

03
scoring:
E-commerce behavior, mobile transactions, and social media activity are transforming
credit assessments, making them more inclusive for unbanked populations.

04 Scalability and local adaptability:

Successful institutions prioritize credit models tailored to local economies, integrating

region-specific factors like rural income patterns or informal sector dynamics.

05 Government support for fintech collaboration:

Programs like Indonesia’s “National Financial Inclusion Strategy” encourage partnerships

between banks, fintechs, and regulators, driving innovation and financial inclusion.

Reimagining Credit with AI 21

5. Three strategic
recommendations for banks
and fintechs to enhance credit
access for the unbanked and
underbanked

01 Adopt localized models:

Why: Credit behaviors and economic activities vary significantly across regions,
particularly in diverse markets like SEA. Using localized data sources enables
institutions to create more accurate and inclusive credit scoring models.
How:
Collaborate with telecom companies to analyze call patterns, mobile payments,
and data usage for credit scoring.
Partner with agricultural cooperatives to incorporate crop yield, market price
trends, and seasonal income data into risk assessments for rural borrowers.
Use region-specific spending data from local e-commerce platforms to refine
credit assessments.

02 Focus on partnerships:

Why: Collaborating with data-rich sectors like telecom, e-commerce, and social media
platforms provides fintechs with alternative datasets to refine credit scoring models.
How:
Partner with telecom companies to analyze mobile usage patterns, bill payments,
and prepaid top-ups.
Collaborate with e-commerce platforms to incorporate purchasing behavior and
transaction histories into risk assessments.
Explore data-sharing agreements with payment gateways and ride-hailing
platforms for behavioral insights.

Reimagining Credit with AI 22

 03 Innovate for scalability:

Why: Underserved markets, like rural communities or gig workers, require solutions
that are both cost-effective and scalable to reach large populations.

How:
Develop AI models that can analyze alternative datasets at scale, ensuring high
accuracy for diverse customer segments.
Design mobile-first platforms that cater to digitally native users with limited
access to formal banking.
Ensure systems are adaptable to varying income patterns, such as seasonal
fluctuations in rural economies or gig worker earnings.

AI-powered credit scoring is reshaping the global financial ecosystem, unlocking

opportunities for millions of underserved individuals. Fintechs and financial institutions
can drive significant impact by leveraging case study insights, fostering innovation, and
collaborating with regulators. This transformative journey offers a roadmap for achieving
greater financial inclusion and sustainable growth in the region.

Reimagining Credit with AI 23

References

1. Bain & Company, e-Conomy SEA 2024 Report.

2. Google and Temasek, SEA Digital Finance Insights 2024.
3. Otoritas Jasa Keuangan (Indonesian Financial Services Authority), Annual Report
2023.
4. World Bank, The State of Financial Inclusion in Southeast Asia.
5. Case Studies from Grab Financial, Bank BRI, and Tonik Bank.
6. https://www.cloudera.com/customers/bank-bri.html
7. https://asianbankingandfinance.net/lending-credit/more-news/indonesias-bank-bris-
mobile-app-cuts-loan-approval-process-2-minutes
8. https://ibsintelligence.com/casestudy/bank-rakyat-indonesia-launches-digital-
lending-apps-powered-by-infosys/
9. https://www.grab.com/sg/inside-grab/stories/from-loans-to-literacy-addressing-
southeast-asias-financing-gap/
10. https://thefintechtimes.com/tonik-joins-up-with-finscore-to-support-underbanked-
filipinos/
11. https://www.apiwiz.io/resources/blogs/case-study--digital-bank-tonik
12. https://kredivocorp.com/2017/11/kredivo-the-first-digital-payment-app-from-
indonesia-to-be-featured-on-google-play-store/
13. https://timo.vn/en/blogs-en/vietnams-timo-bank-builds-with-aws-to-promote-
financial-inclusion/
14. https://www.businesswire.com/news/home/20240430942970/en/Leveraging-AI-
MYbank-Enables-Financing-Services-for-53-Million-SMEs
15. https://www.unionbankph.com/sites/default/files/2023-
04/UnionBank%20Sustainability%20Report%202022.pdf

Reimagining Credit with AI 24