Page 586 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

2 Use Case Description

2�1 Description

In Zambia, 60% of adults remain unbanked due to lack of formal credit histories, geographic
barriers and systemic biases (World Bank, 2020). Mobile money is widespread, yet fraud and
illiteracy hinder its benefits.

The aim is to enable over 500,000 unbanked Zambians to access micro-credit; reduce SMS-
based scam incidence by 60% and deliver financial education.

Innovative Approach include:

Credit Scoring: Explainable AI (logistic regression + Shapley Additive explanations (SHAP) on
mobile transaction patterns (Talk time top-ups, utility payments, etc.). When financial distress is
detected, the system triggers support actions like counselling and tailored credit suggestions
to help the customer regain stability.

Fraud Detection: Real-time anomaly detection at network edge (Lower Latency, Reduced
Backhaul Congestion, Enhanced Privacy etc.)

Chatbots: Localized NLP models (Masakhane) delivering Short Message Service/ Unstructured
Supplementary Service Data (SMS/USSD) guidance in local languages; real-time dispute
resolution via Zambia Information and Communications Technology Authority (ZICTA) hotline
integration for enhanced compliance by service providers.

Use Case Status: The use case is part of a big project in progress

Partners

Zamtel and Zambia Information and Communications Technology Authority(ZICTA)

2�2 Benefits of the use case

Our initiative delivers substantial benefits by expanding access to credit, which significantly
boosts income-generating opportunities for individuals and communities. By providing
microloans, we foster the growth of SMEs and create new job opportunities, driving economic
growth and stability. Leveraging AI on existing mobile infrastructure, we enhance fintech
innovation, making financial services more accessible and efficient. Our fair AI scoring system
ensures that women and customers in rural areas, regardless of their economic status, have
equal access to financial opportunities, thereby reducing inequalities. Proactive fraud detection
mechanisms strengthen trust in institutions by ensuring the integrity and security of financial
transactions. Furthermore, we foster collaboration with operators, regulators, and academia
to create a robust ecosystem that supports sustainable development and inclusive growth.

2�3 Future Work

To further strengthen our platform, we plan to introduce several forward-thinking enhancements.
These include biometric authentication for greater security and ease of use, AI-assisted financial
guidance to help users make smarter financial decisions, and extended dialect support to
make sure the platform is truly accessible and relevant across our region. To enable these

550