NARIT-AI: Gujarat Police’s AI Tool for NDPS Investigations
Why in the news
Gujarat Police rolled out an AI aid for narcotics cases after procedural mistakes kept leading to acquittals and weak conviction numbers.
Key facts
- Tool: NARIT-AI, built by Gujarat Police with AI startup Gradiante Creative Services.
- Problem: NDPS conviction rate fell from 44.4% (2020) to 25% (2022).
- Cause: slips in documentation, witness presence or sample handling lead to benefit-of-doubt acquittals; district offices also lack specialised writers.
- Technology: Retrieval-Augmented Generation (RAG) in a closed knowledge base that avoids invented citations.
- Use: officers upload an FIR and the system gives case-specific guidance.
| Part of knowledge base | Content |
|---|---|
| Statute | NDPS Act |
| New criminal laws | BNS, BNSS, BSA |
| Case law | Thousands of Supreme Court and High Court judgments |
Key concepts
- RAG: the AI consults a given reference database before answering rather than relying on memory that may be outdated.
- Primacy of procedure: NDPS cases need strict step-by-step compliance; evidence alone is rarely enough.
- Role: a paralegal tool that helps police build a strong case file; it does not replace the Public Prosecutor.
Exam angle
- Full form: Narcotics Analysis and RAG-based Investigation Tool.
- Conviction rate numbers: 44.4% to 25%.
- Developer: Gujarat Police with Gradiante Creative Services.