Shared AI infra, compute access should expand public‑interest outcomes: Policy makers

New Delhi: The focus must shift from merely building data‑centre capacity to ensuring resources deliver public‑interest outcomes across health, education and agriculture, senior policy makers said on Friday.

At the ‘India AI Impact Summit 2026’ here, a working report on “Opening Up Computational Resources for New AI Futures,” was launched.

The Ministry of Statistics and Programme Implementation released the report and the session saw speakers stressing "demand aggregation, shared infrastructure, skills development and mission-driven governance frameworks" as critical to translate compute access into real-world deployment for startups, researchers and social-sector organisations.

Senior government leaders, philanthropic institutions and global AI experts examined how catalytic funding, new institutional models and South–South cooperation can make advanced compute accessible and affordable for the Global South.

“We are of the collective opinion that AI will transform the world. The defining question is whether this transformation will be equitable, inclusive, and aligned with the public interest,” said Dr Saurabh Garg, Secretary, Ministry of Statistics and Programme Implementation.

Martin Tisne, CEO, AI Collaborative, said he was optimistic about computing capacity in two years even in the Global South, but doubtful about the effectiveness in use of data centres.

“Transforming AI into a scalable service for consumers and creators is not just a product challenge; it is a policy challenge,” said Vilas Dhar, President, Patrick J. McGovern Foundation.

Dhar underscored the need for new institutional mechanisms to connect policy, capital and deployment at scale, noting that accessibility cannot be left to market forces alone.

Shikoh Gitau, CEO, Qhala, emphasised that compute demand must be anchored in clearly defined development outcomes and supported through cross-country cooperation. “When you have clear use cases, then the GPU demand becomes an obvious task, and the governance framework to bridge these gaps also becomes clearer,” Gitau said.

The session outlined a roadmap of catalytic public and philanthropic capital, shared compute infrastructure and interoperable governance frameworks collectively enabling AI to function as a global public good, the statement noted.

Shared AI infra, compute access should expand public‑interest outcomes: Policy makers