MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641095017 A) filed by St. Peters Engineering College on August 05, 2026, for A Decentralized Blockchain System For Trusted Machine Learning Model Management.
Inventors include Mr. P. Anil Kumar Reddy; Mr. Siddartha Kotha; Mr. S Veeraiah; Dr. P. Ganesh; Dr. S. Satya Nagendra Rao; and Mr. Karumuru Venkat Tiru Gopal Reddy.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention discloses a Decentralized Blockchain System for Trusted Machine Learning Model Management that provides a secure, transparent, and tamper- resistant framework for managing machine learning (ML) models throughout their lifecycle. The system integrates blockchain technology, decentralized storage, smart contracts, cryptographic hashing, and role-based access control to ensure the authenticity, integrity, traceability, and secure sharing of machine learning models across distributed environments. The proposed architecture comprises a machine learning model registration module, blockchain network, decentralized storage layer, smart contract engine, identity and access management module, version control module, integrity verification module, and audit management module. Upon registration, each machine learning model is assigned a unique cryptographic hash that serves as its digital fingerprint. The model files are stored in decentralized storage, while their metadata, ownership information, version history, timestamps, and hash values are immutably recorded on the blockchain. Smart contracts automate model registration, ownership verification, access authorization, deployment approval, version updates, and transaction recording without requiring a centralized authority. Before deployment or sharing, the integrity verification module validates the model by comparing its recalculated hash with the blockchain record, thereby detecting unauthorized modifications or tampering. The system also maintains a permanent audit trail of all model lifecycle events, enabling accountability, provenance tracking, and regulatory compliance. By eliminating single points of failure and providing secure collaborative model management, the invention enhances trust, availability, and transparency in artificial intelligence ecosystems. The proposed system is applicable to healthcare, finance, cybersecurity, manufacturing, Internet of Things (IoT), cloud computing, and other distributed environments where secure and trustworthy machine learning model management is essential.
Disclaimer: Curated by HT Syndication.
