AI-Powered Responsive RTI and IPC Complaint Automation System

Read full text Summarise with VIRAT
Authors: Prabha M, Chandru V, Prakash R, Roghith K, Dharmaraj P Sankalp S
Cited by: 0
References: 22

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The Right to Information Act, 2005, is one of the major building blocks of legislative tools that define the principle of transparency and empowerment of citizens in the Indian polity; but its infallibility in practice often runs afoul of avoiding, half-hearted, or procedurally non-compliant replies by Public Information Officer. Such responses need to be interpreted with particular juridical expertise and the instances of statutory violation be identified, thus obstructing substantive citizen involvement in the accountability mechanism. In order to address this shortcoming, this study suggests an AI-based, data-agnostic RTI intelligence system. The architecture includes a multi-phase pipeline that fuses the implementation of Optical Character Recognition (OCR) and Natural Language Processing (NLP) using transformers and Named Entity Recognition (NER) with an engine based on heuristics and legal inferences to break down the RTI answers. In addition to classification, the framework provides an ordered identification of data requirements and the mechanism of multi-source retrieval prioritizing internal databases, investigating statutory and open-government Application Programming Interfaces such as CPGRAMS, and performing AI-aided cross source-checking to shed light on inconsistencies and disclosure gaps. The generation of drafts to be applied to another application or appeal are posed as a secondary, supported based tool that is only triggered after confirmation of non-compliance. This is to make sure that the outputs of the system also have sound basis on the jurisprudential analysis. The empirical assessment has proven to have strong classification and extraction ability which has been supported by statistical validation using the standard deviation and the confidence interval analysis. These measures testify to the stability of the system in different data sets. The suggested framework leads to strengthening citizen-led public responsibility in digital governance systems by operationalizing structured clarity in the data and AI-aided verification.

Fields of Study & Subjects

Artificial Intelligence in Law E-Government and Public Services Administrative Law and Governance

Publication details

Journal / source
Research Square
Publication date
30 March 2026
Language
EN
Document type
Preprint