MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641085233 A) filed by Cmr Technical Campus; Cmr College Of Engineering & Technology; and Cmr Institute Of Technology on July 11, 2026, for Confidential Federated Learning System With Hardware-Enforced Data Provenance And Verifiable Gradient Integrity.
Inventors include A. Uday Kiran; G. Menaka; Dr. V. Venkataiah; Ms. Komal Parashar; A. Prakash; and Pavan Kumar Panakanti.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: A confidential federated learning system (100) and method are disclosed for enabling secure collaborative model training with hardware-enforced data provenance and verifiable gradient integrity. The system (100) comprises a memory (102), a processor (104), and a communication module (106), operatively associated with a hardware attestation module (108), a data provenance tracking module (110), a gradient integrity verification module (112), a confidential training orchestration module (114), a secure aggregation module (116), a tamper detection module (118), and an audit and compliance module (120). The disclosed system establishes trusted execution environments, verifies provenance of local datasets, generates cryptographic proofs for gradients, and validates authenticity of model updates prior to aggregation. Verified gradients are securely aggregated to update a global model while excluding compromised or unauthorized participant contributions. The invention enhances privacy preservation, trustworthiness, auditability, and resistance against adversarial manipulation in distributed federated learning environments.
Disclaimer: Curated by HT Syndication.
