MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641099391 A) filed by Dr. M Indira Rani on August 17, 2026, for A Deep Learning-Based System For Real-Time Detection Of Injection Attacks In Application Programming Interfaces.
Inventors include Dr. A. Ramesh; B. Raju; Rathod Maheshwar; B. Bhasker; Shashank Reddy Voorelli; Dr. A. Venugopal Rao; Rajitha Tirunagari; and Dr. D. Srinivas.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: APIs are becoming more and more popular for transferring data in web applications and have been a major target for injection attacks, raising substantial issues over system integrity and user privacy. Indeed, the most recent data suggests that injection attacks constitute around 25% of all vulnerabilities in web applications, thus it’s crucial to have good detection tools. However, traditional detection systems only monitor SQL injection assaults and additional attack types such as XML and JSON injections are unmonitored and can be used by hostile actors. To address these issues, this study proposes a novel approach based on the usage of Bidirectional Recurrent Neural Networks (RNNs), to enhance the classification of different types of injection. The work shown that the bidirectional RNNs can extract the best features by studying the data sequences from the front and backward, which improves the accuracy of injection attack detection. The proposed system has better accuracy, precision, and recall than the existing approaches with a simple user interface for real-time predictions. Using this holistic detection solution, organizations may help reduce their vulnerability to injection vulnerabilities and thereby safeguard their apps and user data. Keywords: leaving other forms, such as XML and JSON injections, unmonitored and possibly exploitable by hostile actors
Disclaimer: Curated by HT Syndication.
