MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202611000856 A) filed by Nims University, Jaipur, Rajasthan, on Jan. 5, for 'fraud detection system and method thereof.'
Inventor(s) include Krishna Kant Sharma.
The application for the patent was published on Feb. 13, under issue no. 07/2026.
According to the abstract released by the Intellectual Property India: "The invention provides an AI-powered system and method (100, 200) for real-time detection of fraudulent financial transactions. The system comprises a data ingestion module (101) for acquiring transaction data from multiple sources; a feature extraction engine (102) for deriving static, dynamic, and contextual features; and an AI model ensemble (103) including a deep neural network (DNN) (103A), gradient boosting machine (GBM) (103B), recurrent neural network (RNN) (103C), and anomaly detection model (103D). A real-time scoring engine (104) computes fraud risk scores, while an adaptive learning module (105) updates models continuously. The system further includes a data fusion engine (106), multi-stage verification process (107), explainable AI module (108) with counterfactual reasoning (114), and scalable cloud infrastructure (109). It enables accurate, adaptive fraud detection with reduced false positives and supports federated learning for privacy-preserving collaboration."
Disclaimer: Curated by HT Syndication.
