A Responsible Generative Artificial Intelligence Based Multi-Agent Framework for Preserving Data Utility and Privacy
Open Access
Abstract
The exponential growth in the usage of textual data across industries and data sharing across institutions underscores the critical need for frameworks that effectively balance data utility and privacy. This paper proposes an innovative agentic AI-based framework specifically tailored for textual data, integrating user-driven qualitative inputs, differential privacy, and generative AI methodologies. The framework comprises four interlinked topics: (1) A novel quantitative approach that translates qualitative user inputs, such as textual completeness, relevance, or coherence, into precise, context-aware utility thresholds through semantic embedding and adaptive metric mapping. (2) A differential privacy-driven mechanism optimizing text embedding perturbations, dynamically balancing semantic fidelity against rigorous privacy constraints. (3) An advanced generative AI approach to synthesize and augment textual datasets, preserving semantic coherence while minimizing sensitive information leakage. (4) An adaptable dataset-dependent optimization system that autonomously profiles textual datasets, selects dataset-specific privacy strategies (e.g., anonymization, paraphrasing), and adapts in real-time to evolving privacy and utility requirements. Each topic is operationalized via specialized agentic modules with explicit mathematical formulations and inter-agent coordination, establishing a robust and adaptive solution for modern textual data challenges.Keywords
Funding Information
- Natural Sciences and Engineering Research Council of Canada
- Natural Resources Canada
This publication has 25 references indexed in Scilit:
- Responsible AI Framework for Autonomous Vehicles: Addressing Bias and Fairness RisksIEEE Access, 2025
- Maximizing data utility while preserving privacy through database fragmentationExpert Systems with Applications, 2025
- An Optimized Privacy-Utility Tradeoff Framework for Differentially Private Data Sharing in Blockchain-Based Internet of ThingsIEEE Internet of Things Journal, 2025
- Privacy Utility Tradeoff Between PETs: Differential Privacy and Synthetic DataIEEE Transactions on Computational Social Systems, 2024
- BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning2020
- Membership Inference Attacks Against Machine Learning Models2017
- Differential Privacy2017
- Privacy-preserving smart metering revisitedInternational Journal of Information Security, 2016
- MuDA: Multifunctional data aggregation in privacy-preserving smart grid communicationsPeer-to-Peer Networking and Applications, 2014
- Automata, Languages and Programming2006