Artificial Intelligence for Digital Transaction Fraud Detection and the Credibility of Indonesian Cashless Payments

Purpose: This study analyzes the role of Artificial Intelligence (AI) in improving the effectiveness of fraud detection through real time anomaly detection, behavioral profiling, and machine learning, and evaluates its contribution to the credibility of cashless payment systems in Indonesia.
Research Methodology: A systematic literature review was conducted on national and international scientific articles published between 2020 and 2025 in Scopus, Google Scholar, Garuda, and SINTA, complemented by official reports from Bank Indonesia, the Financial Services Authority, and the Indonesian Fintech Association. The literature was selected using inclusion and exclusion criteria and analyzed thematically. 
Results: The most effective detection approach combines real time anomaly detection for speed, behavioral profiling for misuse of valid credentials, and ensemble machine learning for accuracy, forming a layered defense against increasingly competent fraud perpetrators. Credibility depends not only on algorithmic accuracy but also on ability, integrity through explainability, and benevolence through victim protection. Effective stakeholder synergy is determined mainly by gaps in adoption capacity among institutions.
Conclusions: AI should be positioned as a strategic foundation for the credibility of digital payments, supported by explainable systems and shared data infrastructure.
Liimitations:The study is a literature review and does not empirically test the relationship between AI transparency and user trust. 
Contributions: The study integrates fraud, trust, and diffusion theories to explain how AI strengthens payment credibility and offers recommendations for regulators, payment providers, and researchers.