Participatory AI, Public Sector AI, Differential Privacy, Conversational Interfaces, Explainable AI, Citizen Engagement in AI

2025-04-30

AI Summary

This study addresses the tension between mathematical rigor and democratic accountability in differential privacy (DP) systems for public-sector AI governance. To resolve this, we propose: (1) a novel TOPSIS-based adaptive ε-selection protocol enabling participatory, evidence-informed public deliberation on privacy budgets; (2) an interpretable noise-injection framework integrating real-time MAE visualization and GPT-4–driven impact attribution analysis; and (3) a dynamic regulatory compliance budgeting mechanism that jointly optimizes privacy guarantees, data utility, and legal adherence. Evaluated in a conversational interface system, our approach significantly enhances public understanding of core DP parameters (e.g., ε) and elevates the quality of civic engagement. Crucially, it maintains strict (ε,δ)-differential privacy while shifting AI governance from technocratic opacity toward transparent, democratically coordinated decision-making.

Technology Category

Machine Learning: PrivacyNatural Language Processing: Ethics — Bias, Fairness, Transparency & PrivacyPhilosophy and Ethics of AI: Privacy & Security

Application Category

Security and Privacy: Data transparency and provenanceUser Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalizationEconomics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments

Abstract

This paper introduces a conversational interface system that enables participatory design of differentially private AI systems in public sector applications. Addressing the challenge of balancing mathematical privacy guarantees with democratic accountability, we propose three key contributions: (1) an adaptive $epsilon$-selection protocol leveraging TOPSIS multi-criteria decision analysis to align citizen preferences with differential privacy (DP) parameters, (2) an explainable noise-injection framework featuring real-time Mean Absolute Error (MAE) visualizations and GPT-4-powered impact analysis, and (3) an integrated legal-compliance mechanism that dynamically modulates privacy budgets based on evolving regulatory constraints. Our results advance participatory AI practices by demonstrating how conversational interfaces can enhance public engagement in algorithmic privacy mechanisms, ensuring that privacy-preserving AI in public sector governance remains both mathematically robust and democratically accountable.

Problem

• Balancing mathematical privacy with democratic accountability in public AI
• Designing participatory AI systems with differential privacy parameters
• Ensuring privacy-preserving AI complies with dynamic legal constraints

Innovation

• Adaptive ε-selection protocol using TOPSIS for DP
• Explainable noise-injection with MAE and GPT-4
• Dynamic privacy budget based on legal compliance