The Race — Building Democracy in the Age of AI. People race toward a shared finish line.
The Race
← Specific Events

Delhi, India

Scorecard

Delhi launches Data DHARA AI-ready public data architecture — August 25, 2026

What happened

Delhi’s Planning Department signed an MoU with EkStep Foundation to prepare government administrative data for AI applications and develop AI-enabled public-service use cases. At the accompanying government-wide workshop, Delhi launched the Data DHARA toolkit and vision paper. DHARA is an open-source, digital-public-infrastructure-based methodology designed to harmonize and prepare administrative data for AI use, with pluggable, interoperable, cloud and on-premises deployment options.

Potential direction of power

Dispersion · Dispersion +1

Delhi is investing in a reusable public data substrate rather than relying solely on proprietary AI applications purchased department by department. That potentially strengthens government implementation capacity while preserving more choice over vendors, models and deployment environments. The score remains +1 because the practical degree of interoperability and institutional control will depend on implementation.

AI implementation capacity increasingly depends on the infrastructure beneath the model itself. Harmonized, reusable government data can become a form of public/civic infrastructure, allowing agencies to build multiple applications on a common foundation and potentially reducing fragmentation and vendor dependence. Delhi therefore provides a useful example of governments building the prerequisites for AI capacity rather than merely buying access to AI systems.

Window of contestability

Opens/Preserves

Positive, with important unresolved governance questions. Open-source, interoperable and on-premises deployment options reduce architectural dependence on any single vendor and preserve greater institutional control over the underlying data layer. Public information is still limited, however, on data-access rules, privacy and consent, auditability, redress, procurement arrangements, and governance of cross-departmental data integration.

Supporting sources

  1. linkedin.com · Supporting source 1 ↗