MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641095981 A) filed by Divya Pachimatla; and Dr. M. Janga Reddy on August 08, 2026, for System And Method For Verifiable Data Lineage And Model Attribution Using Cryptographically Linked Provenance Graphs.
Inventors include P. Deekshith Chary; P. Kumar; S. Sreshta Joshi; S. Parimala; Gourishetty Raga Mounika; and Thejovathi.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: A system (100) and method for verifiable data lineage tracking and model output attribution using cryptographically linked provenance graphs are disclosed. The system (100) comprises a memory (102), a processor (104), and a communication module (106) configured to receive source datasets, transformation metadata, workflow execution events, and model-related information associated with one or more computational environments. A provenance graph construction module (108) generates provenance graphs representing data entities, transformation operations, model states, and output relationships. A cryptographic linkage module (110) establishes tamper-resistant hash associations between provenance nodes and relational edges. A constraint validation module (114) validates provenance integrity against predefined governance and consistency constraints, while an attribution determination module (116) associates generated outputs with originating datasets and transformation paths. A distributed synchronization module (118) maintains lineage consistency across distributed nodes. The disclosed framework enables secure, verifiable, and auditable end-to-end computational traceability for artificial intelligence and data processing systems. Fig. 1
Disclaimer: Curated by HT Syndication.
