MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641072768 A) filed by Aditya Institute Of Technology And Management on June 11, 2026, for A Self-Evolving Artificial Intelligence Method For Human-Centric Explainable Intelligence And Trust-Aware Automation.
Inventors include Mrs. Ratnamohitha Maddula; Mrs. Kopara Madhuri; Mrs. Modalavalasa Divya; Mrs. Modalavalasa Varalakshmi; Mrs. Bora Srujana; Mrs. Palleda Sravani; Mrs. Lakkaji Mounika; Mrs. Thamira Pooja; Miss. Bhavani Manchala; and Ms. Nadiminti Shailaja.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: The present invention discloses a self-evolving artificial intelligence method for human-centric, explainable, and trust-aware automation across multiple application domains. The method integrates contextual data acquisition, adaptive decision generation, automated explanation construction, and quantitative trust evaluation within a unified operational framework. For each generated decision, the system produces a structured, human-understandable explanation identifying contributing factors, reasoning pathways, and confidence indicators. A trust assessment module computes reliability metrics including uncertainty levels, fairness indicators, bias detection parameters, and regulatory compliance alignment prior to output deployment. If trust thresholds are not satisfied, corrective reasoning or human oversight may be initiated. The method further incorporates continuous self-evolution through feedback-driven learning, reinforcement updates, and outcome validation to refine both decision and explanation models over time. The invention enhances transparency, accountability, and user confidence while maintaining high decision accuracy, enabling responsible AI deployment in healthcare, finance, education, governance, and industrial automation systems.
Disclaimer: Curated by HT Syndication.
