MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641062186 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on May 15, for 'a system for trust-based autonomous dataset validation and integrity assurance in artificial intelligence.'
Inventor(s) include Dr. Santhi K; and Rayban Pranav Mahesh.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a trust-aware system and method for constructing and validating datasets for artificial intelligence applications. The system comprises a multi-layer architecture including data ingestion, multi-factor trust scoring, bias analysis, adversarial detection, cryptographic integrity verification, adaptive weight assignment, and dataset generation. Each data sample is evaluated based on parameters including source authenticity, statistical conformity, anomaly absence, temporal freshness, and contamination likelihood. The system employs an adaptive threshold mechanism based on model performance to refine validation criteria. Adversarial data is detected using an ensemble of techniques, and bias is evaluated using fairness metrics. Cryptographic fingerprinting ensures data integrity and auditability. The system further assigns adaptive weights to improve training efficiency. The output is a validated dataset with integrity certification and metadata, enabling secure, reliable, and fair AI model training."
Disclaimer: Curated by HT Syndication.
