AI FOR WELFARE ACCESS: A SYSTEMATIC REVIEW OF FRAMEWORKS FOR ENHANCING ACCESSIBILITY AND REACHABILITY OF GOVERNMENT SCHEMES

Abstract

Government welfare delivery in India increasingly relies on digital portals and online systems; however, effective access to schemes remains uneven due to digital divides, fragmented information, and limited accessibility. Many citizens, particularly from rural and marginalized communities, struggle to identify relevant schemes, understand eligibility conditions, and complete application processes. Recent policy initiatives highlight the potential of artificial intelligence (AI) to enhance public-service delivery, yet existing AI- based welfare systems remain portal-centric, weakly explainable, and insufficiently inclusive.
This paper proposes a design-science–oriented, AI- enabled conceptual framework to improve the accessibility and reachability of government welfare schemes. The framework integrates rule-based eligibility assessment with learning-based prioritization, multimodal and multilingual conversational interfaces, explainable AI mechanisms, and human-in-the-loop participation by frontline workers. A prototype-level feasibility demonstration using a simulated use-case scenario is presented to illustrate operational viability. The proposed framework provides a structured foundation for future pilot implementations and empirical evaluation and contributes toward the development of inclusive, transparent, and citizen- centric AI systems aligned with Digital India and the AIforAll strategy.