Government has introduced a Model Context Protocol server to facilitate linking of Artificial Intelligence tools with Official Statistical databases

1. At a Glance

MCP server launched by the National Statistics Office (NSO) under the Ministry of Statistics & Programme Implementation (MoSPI) to bridge AI tools (Claude, ChatGPT, Cursor, etc.) with India's official statistical databases via the eSankhyiki portal [S1][S2][S3].
Allows natural-language querying of verified government statistics without manual PDF/CSV downloads — converts static datasets into AI-consumable APIs [S2][S4].
UPSC relevance: intersects GS-II (governance, e-governance, transparency) and GS-III (Sci-Tech, data, AI); classic factual hook for Prelims (ministry, products, portal name) [S1].

23 March 2026: MoS (I/C) for Statistics & PI informed Parliament that the beta MCP server, which initially covered 7 key statistical products, has been scaled up to 21 products on eSankhyiki [S1].
6 February 2026: NSO originally launched the beta MCP server to make government data "AI-ready" [S2][S3].

2024: MoSPI launched the eSankhyiki portal as a one-stop platform for official statistics, replacing fragmented PDF dissemination with API-based access [S5].
eSankhyiki's API layer feeds the National Data and Analytics Platform (NDAP) of NITI Aayog [S5].
Feb 2026: Beta MCP server released as a pilot project (source code on GitHub under MIT licence by nso-india/esankhyiki-mcp) [S3].
Mar 2026: Expansion from 7 → 21 datasets announced via PIB [S1].

Parent ministry: Ministry of Statistics & Programme Implementation (MoSPI) [S1].
Implementing body: National Statistics Office (NSO) [S2].
Host portal: eSankhyiki (esankhyiki.mospi.gov.in) [S5].
Technology: Model Context Protocol (MCP) — open standard linking LLM/AI agents to data sources [S2][S3].
Licence: MIT (open source on GitHub) [S3].
Beta — 7 products: Periodic Labour Force Survey (PLFS), Consumer Price Index (CPI), Annual Survey of Industries (ASI), Index of Industrial Production (IIP), National Accounts Statistics (NAS), Wholesale Price Index (WPI), Environmental Statistics [S2].
Extended — 21 products include the above plus datasets on GDP, energy, higher & school education, gender, health, trade, agriculture, consumption, economic census, digital literacy [S3].
Downstream linkage: Feeds NDAP (NITI Aayog) via API [S5].

Scientific / Technological
MCP is an open client-server protocol that standardises how AI assistants access tool/data context — India is among early State adopters for official statistics [S2][S3].
Removes brittle scraping; provides structured, schema-aware endpoints over verified data [S2].
Governance / Administrative
Strengthens data-driven governance by giving policymakers real-time, queryable access [S2].
Reduces information asymmetry between Centre, States, researchers, and citizens [S2].
Economic
Lowers transaction cost of using official data for businesses, startups, and researchers; supports evidence-based decision-making [S2].
Supports indicators central to RBI/MPC (CPI, IIP, WPI) being usable by AI analytics stacks [S2].
Ethical / Transparency
Anchors AI outputs to verified government data, mitigating hallucination and misinformation on Indian macro indicators [S2].
Open-source release (MIT) signals transparency and invites civic-tech contribution [S3].
Federalism
Statistics is in the Union List (Entry 94) and Concurrent List (Entry 45); MoSPI's centralised AI gateway raises questions of State-level statistical integration (relevant context, not in source).

6 Feb 2026: NSO launches beta MCP server with 7 datasets [S2].
Feb 2026: Source code published on GitHub (nso-india/esankhyiki-mcp) under MIT [S3].
23 Mar 2026: Parliamentary statement confirms expansion to 21 statistical products [S1].

MCP server for official statistics launched by NSO under MoSPI — not MeitY, not NITI Aayog [S1].
Hosted on the eSankhyiki portal [S1].
Beta version: 7 products; extended: 21 products [S1].
Beta 7 = PLFS, CPI, ASI, IIP, NAS, WPI, Environmental Statistics [S2].
PLFS is conducted by NSO under MoSPI [S2].
CPI released by NSO; WPI released by Office of Economic Adviser, DPIIT (trap — only CPI on MCP is MoSPI's) [S2].
NDAP (which consumes eSankhyiki APIs) is a NITI Aayog platform [S5].
MCP source code released under MIT licence on GitHub [S3].
Beta MCP server went live on 6 February 2026 [S2].
MCP = Model Context Protocol, an open standard for connecting AI tools to data/tool servers [S2].

GS-II: e-Governance — applications, models, successes, limitations; transparency & accountability.
GS-III: Awareness in IT, Computers; achievements of Indians in S&T; indigenisation of technology.
Plausible question stems: 1. "Linking AI tools with verified official statistics can transform evidence-based policymaking in India. Discuss with reference to the MoSPI MCP initiative." (GS-II/III, 15M) 2. "How can open protocols like the Model Context Protocol mitigate AI hallucination risks in public administration?" (GS-III, 10M) 3. "Discuss the role of platforms such as eSankhyiki and NDAP in democratising official data." (GS-II, 10M)

eSankhyiki Portal — direct parent platform for the MCP server [S5].
National Data and Analytics Platform (NDAP) — NITI Aayog's data aggregator that consumes eSankhyiki APIs [S5].
Periodic Labour Force Survey (PLFS) — flagship MoSPI survey on the MCP [S2].
Consumer Price Index (CPI) & IIP — high-frequency MoSPI indicators on the MCP [S2].
IndiaAI Mission (MeitY, 2024) — broader AI ecosystem context.
DPDP Act, 2023 — data-protection regime governing such platforms.
National Statistical Commission (NSC) — apex advisory body for official statistics.
Open Government Data (OGD) Platform — data.gov.in — predecessor open-data initiative.

Confusing MCP server (open AI–data protocol) with a generic "government cloud server" or with India's AI compute mission.
Attributing the launch to MeitY or NITI Aayog; it is MoSPI/NSO.
Confusing eSankhyiki (MoSPI dissemination portal) with NDAP (NITI Aayog analytics platform) — they are linked but distinct [S5].
Assuming WPI is a MoSPI product — it is released by DPIIT; only its inclusion on the MCP server is via the eSankhyiki API.
Mis-stating beta coverage (it is 7, expanded to 21, not 21 from day one) [S1].

[S1] Government has introduced a Model Context Protocol server… (PIB, MoSPI, 23 Mar 2026) — https://www.pib.gov.in/PressReleasePage.aspx?PRID=2243781 — (tier: 1)
[S2] NSO Launches MCP Server to Make Government Data AI Ready (PIB) — https://www.pib.gov.in/PressReleasePage.aspx?PRID=2224472 — (tier: 1)
[S3] eSankhyiki MCP pilot source code (NSO India, GitHub) — https://github.com/nso-india/esankhyiki-mcp — (tier: 1, government repo)
[S4] MoSPI MCP Beta — Taking AI to Data — https://datainnovation.mospi.gov.in/mospi-mcp — (tier: 1)
[S5] MoSPI Launches eSankhyiki Portal (PIB) — https://www.pib.gov.in/PressReleasePage.aspx?PRID=2029708 — (tier: 1)