MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641104729 A) filed by Vardhaman College Of Engineering on August 31, 2026, for Adaptive Ethical Reasoning In Autonomous Al Systems Using Reinforcement Learning.
Inventors include Ms. Sumaiya S. K.; Ms. Ch. Dhanalaxmi; Dr. K. Mamatha; Dr. Vinayak G Biradar; Ms. Shruti S Soma; and Ms. B. Swapna.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: Adaptive Ethical Reasoning in Autonomous Al Systems Using Reinforcement Learning is the proposed invention. The proposed invention integrates contextual data acquisition, adaptive ethical knowledge management, and a PPO-based actor-critic reinforcement learning model to learn optimal policies balancing operational efficiency and ethical compliance. The multi-objective reward function considers the actions in terms of safety, fairness, privacy, transparency, accountability and compliance with regulations, allowing the agent to improve the decision-making process constantly by interacting with the environment. The system also includes an explainable AI module that provides transparent reasoning for each autonomous decision, and a human-in-the-loop feedback mechanism that refines ethical policies with expert guidance. The framework can adjust to changing regulations, environmental conditions, and societal expectations through continuous monitoring and incremental updates to policies, while retaining previously learned knowledge. The invention proposed improves the reliability, transparency, adaptability, and trustworthiness of autonomous AI systems. It is suitable for applications such as autonomous vehicles, healthcare, intelligent robotics, industrial automation, financial services, and smart city infrastructure.
Disclaimer: Curated by HT Syndication.
