The People’s AI Watchdog: A Citizen-Led Platform for Transparent and Accountable AI Governance

Abstract

The rapid integration of Artificial Intelligence into public-sector decision-making has improved efficiency but has also introduced challenges related to transparency and accountability. Citizens often receive automated outcomes without clear explanations, making it difficult to understand or question these decisions. This work presents a system-focused solution in the form of a Citizen-Led AI Policy Audit Platform, designed to capture, process, and visualize real-world experiences of individuals affected by AI-driven systems.The platform allows users to submit their experiences through a simple interface, where the data is securely processed using anonymization techniques and Natural Language Processing methods. The system extracts key metadata from user narratives and stores it in a structured format for further analysis. An interactive dashboard is then used to visualize trends, patterns, and potential biases across different domains.The implementation demonstrates a complete workflow from data collection to visualization, supported by user interfaces and administrative dashboards. Through this approach, the system transforms individual experiences into meaningful insights, enabling better understanding of AI-driven decision patterns. The platform emphasizes usability, privacy, and transparency, providing a practical framework for participatory AI auditing in governance systems.