MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641106246 A) filed by Uma R; Hemalatha J; R. Sangeetha; Durai Chinnappan J; Dr. C. Thangamani; C. Subbulakshmi; Dr. B. Rethina Kumar; and Dr. C. S. Gayathri on September 03, 2026, for Artificial Intelligence-Based System And Method For Automated Detection And Verification Of Misinformation In Digital Content.
Inventors include Uma R; Hemalatha J; R. Sangeetha; Durai Chinnappan J; Dr. C. Thangamani; C. Subbulakshmi; Dr. B. Rethina Kumar; and Dr. C. S. Gayathri.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: The present invention provides an artificial intelligence-based system and method for automated detection and verification of misinformation in digital content. The system receives textual, image, audio, video, web-based or multimodal digital content and preprocesses the received content for subsequent analysis. A multimodal feature extraction module extracts semantic, linguistic, visual, acoustic, temporal and metadata features, while a claim extraction module identifies factual claims. An artificial intelligence analysis module analyses the extracted claims and an evidence retrieval module retrieves corresponding information from configured tr- usted repositories. A source credibility module evaluates the reliability of the content source, while a verification engine compares the extracted claims with retrieved evidence and determines evidence agreement. A multimodal consistency module evaluates consistency between different content modalities. A misinformation assessment module generates a misinformation probability and confidence score based on claim analysis, evidence agreement, source credibility, contextual validity and multimodal consistency. The system generates an automated classification and an explainable verification report identifying supporting evidence, contradictory evidence and detected inconsistencies.
Disclaimer: Curated by HT Syndication.
