Push for more AI funds: Industry wants IndiaAI Mission outlay doubled

MeiTY, according to a senior govt official, has received no such proposal and it is not considering any such move

AI infrastructure providers and deeptech players have sought a doubling of the outlay for the IndiaAI Mission, with an expanded scope to support more ambitious projects, sources aware of the development said.

 

The additional funding will be required for onboarding and subsidising additional Graphics Processing Units (GPUs), as well as bringing more startups and companies under the IndiaAI Mission umbrella to build indigenous foundational models, Large Language Models, Large Reasoning Models, and Small Language Models, industry sources said.

 

“They (government) are willing to put in additional budgets. They want to double it but in blocks and talks for that are ongoing,” one of the industry persons, mentioned on condition of anonymity.

 

Another industry player in the GPU and cloud services segment also added that though the idea is to eventually double the corpus under the mission, "the exact timelines or the amount of funds to be allocated have not been decided yet, " he added. 

 

The Ministry of Electronics and Information Technology has, however, received no such proposal from the industry, and it is not considering any such move, a senior government official said. Officials said that there is still a substantial amount of the outlay of the ₹10,372 crore remaining to be deployed.

 

Last year, the Cabinet allocated approximately ₹10,372 crore under the Mission for five years to catalyse various components, including pivotal initiatives such as IndiaAI Compute Capacity, IndiaAI Innovation Centre, IndiaAI Datasets Platform, IndiaAI Application Development Initiative, IndiaAI FutureSkills, IndiaAI Startup Financing, and Safe & Trusted AI.

 

The aim was to nurture the country’s AI innovation ecosystem through a public-private partnership model. So far, India has selected four startups — Sarvam, SoketAI, Gan AI, and Gnani AI — to develop indigenous foundation models as it seeks to gain a foothold in sovereign AI.

 

Sarvam AI has been tasked with building the country’s first sovereign LLM ecosystem, developing an open-source 120-billion-parameter AI model to enhance governance and public service access through use cases like “2047: Citizen Connect and AI4Pragati”.

 

SoketAI will develop India’s first open-source 120-billion-parameter foundation model optimised for the country’s linguistic diversity, targeting sectors such as defence, healthcare, and education.

 

Gnani AI will build a 14-billion-parameter voice AI foundation model delivering multilingual, real-time speech processing with advanced reasoning capabilities, while Gan AI will create a 70-billion-parameter multilingual foundation model targeting text-to-speech capabilities.

 

India expanded its GPU capacity to nearly 40,000 from 18,417 earlier this year, Union Minister for Electronics and Information Technology Ashwini Vaishnaw had said.

 

This expanded Cloud compute capacity is expected to provide a common computational AI platform for training and inference, which is crucial for developing indigenous foundation models and AI solutions tailored to the Indian context.

 

Considering the geopolitical context, there is a consensus that India should strive to become more independent across every layer of the AI stack, one of the sources said, responding to the Mission goals.

 

The industry demand to expand the corpus comes even as the government is in the process of awarding contracts to build LLMs to eight more companies, including Tech Mahindra, Fractal, BharatGen, Genloop, and four other enterprises, one of the people quoted above said.