Cabinet approves ₹10,372-crore for India AI Mission

The Union cabinet has approved a ₹10,372-crore India AI Mission that will provide funding for deep-tech AI startups as well as seek to develop an open-source database that can be used to train artificial intelligence models and other applications.

The mission will also provide a framework for developing graphic processing units (GPUs) under public-private partnerships as well as multi-model, domain-specific large language models that power generative AI platforms.

“A multi-faceted comprehensive framework has been created to bolster an AI-based solutions ecosystem in India,” Union minister of commerce Piyush Goyal said at a briefing following the cabinet approval on Thursday.

“The Mission listed seven key objectives—building compute capacity…; innovation centres to build LLMs that will cater to singular industries; a datasets platform to offer non-personal data for all interested parties; an ‘application development initiative’ that will create a marketplace for AI services that will be commercially marketed; a ‘FutureSkills’ initiative that will create AI courses in undergraduate, postgraduate and research divisions; a startup funding initiative 

that will “support and accelerate deep-tech AI startups and provide them streamlined access to funding to enable futuristic AI projects,” Goyal said.

It will also recognize “the need for adequate guardrails to advance the responsible development, deployment, and adoption of AI”—in order to build what the government refers to as ‘safe and trusted AI’. 

The government will create two or three innovation centres to develop domain-specific LLMs, Goyal said. 

He added that the unified data platform under the AI Mission will be a “one-stop solution” towards making data available to startups and researchers. 

It was not immediately clear if Bhashini, the ministry of electronics and information technology (Meity)’s Indic language datasets project, will continue to operate as an independent entity or be merged into the unified platform. 

In October, Rajeev Chandrasekhar, Union minister of state for IT, stated that seven working committees had submitted a report on creating a framework for the India AI Mission. 

The report highlighted a need to create a public-private partnership model to develop indigenous compute capacity—which academia and researchers had long clamoured for.

“While the quantum may seem small by itself, the PPP model is likely to see large tech conglomerates such as Google or Microsoft join hands with a Centre-backed entity to create the compute infrastructure,” Jayanth Kolla, partner at industry consultant Convergence Catalyst, told Mint.

“Further, the development of the domain-specific AI models will cater to specific usage requirements of each ministry—such as a health-specific LLM catering to communications and data offerings of the Union health ministry,” Kolla said.

Arun Prabhu, partner at law firm Cyril Amarchand & Mangaldas, added that the formal mission can offer “a great opportunity for innovators to partner with the government, and enable accelerated development.”