Publicação
Artificial Intelligence and Blockchain in Digital Banking: Automating Loan Management from Origination to Fraud Prevention
| Resumo: | This thesis investigates the integration of Artificial Intelligence (AI) and Blockchain technology to fully automate and enhance loan management processes within digital banking. Traditional lending processes are often hindered by manual documentation verification, slow approval times, limited transparency, and vulnerability to fraudulent activities. To address these challenges, this study developed a comprehensive digital lending framework leveraging AIpowered chatbot interactions, automated document retrieval and validation, dynamic creditworthiness evaluations based on behavioral insights from bank statements, and advanced fraud detection mechanisms using machine learning models. Additionally, Blockchain technology was employed through smart contracts to ensure secure, transparent, and immutable recording of loan transactions. The methodology involved developing specialized AI agents using large language models, training a fraud detection model on synthetic financial data, and building a Minimum Viable Product (MVP) to simulate end-toend loan applications. Results demonstrated remarkable efficiency improvements, achieving a 99.98% reduction in document processing times compared to traditional methods, alongside enhanced accuracy in credit evaluations and fraud detection. The integrated system provided personalized, transparent, and real-time interactions, significantly improving customer experience and trust in automated banking solutions. While practical implementation highlighted challenges concerning data privacy, AI model security, and potential manipulation risks, these issues were addressed through strategic recommendations for banks, technology providers, and policymakers. The study concludes that the combined application of AI and Blockchain technologies represents a substantial advancement in digital banking, promising to deliver greater operational efficiency, robust security, and improved user satisfaction. Future research directions include further optimization of AI models and expanded empirical testing within real-world banking environments. |
|---|---|
| Autores principais: | Ferreira, Pedro Cerejeira Príncipe |
| Assunto: | Large Language Models (LLMs) Blockchain Bank Automation Data Security Fraud Prevention SDG 8 - Decent work and economic growth SDG 9 - Industry, innovation and infrastructure SDG 10 - Reduced inequalities SDG 16 - Peace, justice and strong institutions SDG 17 - Partnerships for the goals |
| Ano: | 2025 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso embargado |
| Instituição associada: | Universidade Nova de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório Institucional da UNL |
| Resumo: | This thesis investigates the integration of Artificial Intelligence (AI) and Blockchain technology to fully automate and enhance loan management processes within digital banking. Traditional lending processes are often hindered by manual documentation verification, slow approval times, limited transparency, and vulnerability to fraudulent activities. To address these challenges, this study developed a comprehensive digital lending framework leveraging AIpowered chatbot interactions, automated document retrieval and validation, dynamic creditworthiness evaluations based on behavioral insights from bank statements, and advanced fraud detection mechanisms using machine learning models. Additionally, Blockchain technology was employed through smart contracts to ensure secure, transparent, and immutable recording of loan transactions. The methodology involved developing specialized AI agents using large language models, training a fraud detection model on synthetic financial data, and building a Minimum Viable Product (MVP) to simulate end-toend loan applications. Results demonstrated remarkable efficiency improvements, achieving a 99.98% reduction in document processing times compared to traditional methods, alongside enhanced accuracy in credit evaluations and fraud detection. The integrated system provided personalized, transparent, and real-time interactions, significantly improving customer experience and trust in automated banking solutions. While practical implementation highlighted challenges concerning data privacy, AI model security, and potential manipulation risks, these issues were addressed through strategic recommendations for banks, technology providers, and policymakers. The study concludes that the combined application of AI and Blockchain technologies represents a substantial advancement in digital banking, promising to deliver greater operational efficiency, robust security, and improved user satisfaction. Future research directions include further optimization of AI models and expanded empirical testing within real-world banking environments. |
|---|