Real-Time Preemptive Fraud Detection Using Adaptive Agentic AI for Financial Transaction Risk Intelligence
International Journal of Research and Applied Innovations · 22.09.2025 · 8 (5)
DOI: 10.15662/ijrai.2025.0805028 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Research and Applied Innovations, online erschienen am 22.09.2025. Verfasst von Manas Ranjan Panda.
Abstract (Originalsprache)
The sudden change of digital financial services has revolutionized the financial transactions, made them complicated, fast and large volume making it so challenging to the traditional financial fraud detection paradigms, which relies on predetermined patterns and a sluggish process. A Real-Time Preemptive Fraud Detection Framework is proposed in this study based on the Adaptive Agentic Artificial Intelligence (RPAFAI) to determine the intelligent financial transactions risk. The suggested system combines real-time transaction information, behavior categorization, situational risk assessment, dynamic machine learning, and self-sufficing AI programs to detect possible fraudulent actions prior to the conclusion of the transaction. The framework is further subdivided into six key levels, namely: (1) Transaction Data Ingestion Layer that constantly gathers transactional and contextual data; (2) Feature and Behavioral Intelligence Layer on which time and transactional-based requirements are temporally processed along with user/behavioral requirements, which are provisioned into future detection results; (3) Adaptive Agentic AI Layer that temporarily processes transactional capabilities in which the special AI agent gathers and an incorporated output that is translated into the future detection results; (4) Real-Time Risk Intelligence Layer where time and transactional-based requirements are processed continuously, alongside user/behavioral requirements, which are incorporated into future detection results; (5) Preemptive Decision and Response Layer that continually incorporates visible results into future detection results; and (6) Feedback and Continuous Learning Layer that constantly integrates results of known outcomes into future doubts. The framework is designed to minimize detection time, false positive and maximize its use to new trends of frauds and maximize real time financial risks knowledge. The proposed methodology provides an arena of active and independent prevention of frauds in the modern digital financial arena and ecosystems.
Zitieren
APA
Panda, M. R. (2025). Real-Time Preemptive Fraud Detection Using Adaptive Agentic AI for Financial Transaction Risk Intelligence. International Journal of Research and Applied Innovations, 8(5). https://doi.org/10.15662/ijrai.2025.0805028
Vancouver
Panda MR. Real-Time Preemptive Fraud Detection Using Adaptive Agentic AI for Financial Transaction Risk Intelligence. International Journal of Research and Applied Innovations. 2025;8(5). doi:10.15662/ijrai.2025.0805028
BibTeX
@article{Panda2025,
author = {Panda, Manas Ranjan},
title = {Real-Time Preemptive Fraud Detection Using Adaptive Agentic AI for Financial Transaction Risk Intelligence},
journal = {International Journal of Research and Applied Innovations},
year = {2025},
volume = {8},
number = {5},
doi = {10.15662/ijrai.2025.0805028}
}
EUVIMED