MNT-Halan's AI Alternative Credit Scoring Engine Automates Over Half of Loan Approvals

MNT-Halan, an Egyptian fintech company and the Middle East and North Africa's first fintech unicorn, has developed an AI-powered alternative credit scoring engine that evaluates users from behavioural and transactional data, many of whom have never used formal credit, according to an October 2025 article on the World Economic Forum's site by co-founder and CTO Ahmed Mohsen. With this proprietary model, MNT-Halan has automated more than 50% of its loan approvals and achieved a 60% approval rate for previously unscoreable users. The engine draws on data from the Halan super-app, which offers payments, savings, investment and e-commerce services. Users' purchases, payment history and interactions with the app feed directly into the model, which builds dynamic, individual profiles. In one example journey, an unbanked customer buys wholesale groceries through the app's e-commerce service, builds a score, receives a consumer finance limit, then the Halan Card, and later secured investment products. MNT-Halan applies five responsible-AI principles: fairness; explainability through transparent scorecards; privacy, with user consent to data sharing and compliance with Egypt's personal data protection law; ongoing monitoring; and human oversight, with sensitive or ambiguous cases reviewed manually. Having started operations in Turkey, Pakistan and the United Arab Emirates, the company is working to expand access to credit across the region.