For much of the past few years, discussions about artificial intelligence (AI) in India have revolved around chatbots, virtual assistants, and applications built on large language models. But as companies move from experimenting with AI to deploying it at scale, focus is increasingly shifting to the infrastructure that powers the technology, including data centres, computing capacity, cloud infrastructure, connectivity and energy.
This shift was visible at Reliance Industries’ annual general meeting (AGM) 2026, where the company outlined plans to build what it calls a “sovereign AI backbone” in India, backed by data-centre infrastructure, graphics processing units (GPUs), renewable energy and partnerships with global technology companies.
As AI adoption gathers pace, the next chapter of India’s AI story may be shaped as much by investments in compute and connectivity as by breakthroughs in AI models and applications.
Why compute has become the new battleground
During the first wave of AI adoption, driven by applications, companies experimented with chatbots, coding assistants, search tools, and content-generation platforms. The next phase is set to be defined by the infrastructure needed to run those services efficiently and at scale.
“The first wave of AI adoption was driven by applications because they were the most visible manifestation of the technology. However, as enterprises move from experimentation to deployment at scale, the underlying infrastructure becomes the critical constraint,” Sunil Kharbanda, founder and chief operating officer at Trezix, a Surat-based AI-led global trade platform, told Business Standard.
Reliance’s AI blueprint
Reliance Industries said it is building a sovereign AI backbone in Jamnagar that will be powered entirely by clean energy generated from its renewable energy assets. The first phase, comprising 120 megawatts of capacity, is expected to be commissioned by the end of 2026.
It also disclosed plans to operationalise an initial fleet of Nvidia GB300 GPUs. According to the company, this is equivalent to more than 75,000 Nvidia H100 GPUs on an AI inference basis. Once the first phase becomes fully operational, the capacity could scale to more than 200,000 H100-equivalent GPUs.
The company has also expanded partnerships with Google, Meta and Nvidia. While Google AI Pro is being offered to Jio users, the Meta partnership is focused on enterprise AI applications and model deployment.
Kharbanda said compute capacity, GPUs, data centres, cloud platforms, networking, and storage are increasingly becoming strategic assets. “Applications remain important because they create demand and business value, but infrastructure is what ultimately determines how much AI activity a country can support and how quickly it can innovate,” he said.
A wider industry push
Reliance is not alone in increasing its focus on AI infrastructure. Several companies in India are investing in different parts of the ecosystem.
For example, Adani Group has been expanding its presence in data centres while also investing in renewable energy and transmission infrastructure.
Bharti Airtel comes closer on the connectivity front, as with its telecom operations, the company has been growing its Nxtra data-centre business and strengthening its enterprise offerings.
The Tata Group, through Tata Consultancy Services (TCS), has invested in cloud partnerships.
Then there are specialist players such as Yotta and CtrlS, which have focused largely on building data centres and computing infrastructure. These firms are focused on a narrower slice of the value chain.
Jaspreet Bindra, founder of AI advisory firm AI&Beyond India and Tech Whisperer Limited UK, said recent global developments have highlighted the importance of having domestic compute and data infrastructure, particularly for strategic resilience and sensitive sectors. India will eventually need its own compute capacity, data infrastructure and possibly even indigenous large language models, he said.
Ritwik Batabyal, chief technology and innovation officer at Mastek, told Business Standard that sovereign AI infrastructure is becoming increasingly important as AI gets embedded in sectors such as healthcare, financial services, manufacturing and public administration. However, he said India’s AI ecosystem would benefit most from a balanced approach that combines global technology partnerships with investments in local infrastructure, talent and innovation.
The building blocks of AI
The next stage of India’s AI journey may be determined less by who builds the most popular chatbot and more by who can provide affordable access to computing power.
“The technology itself is advancing faster than organisational transformation,” said Kharbanda. Beyond access to compute, companies are grappling with issues such as data integration, governance, cybersecurity, and identifying AI use cases that deliver clear returns.
The experts highlighted that over the next three to five years, the biggest AI investments in India are likely to occur in data centres, cloud infrastructure and AI compute because they form the foundation for everything else. Significant capital will be required to build GPU clusters, networking infrastructure, storage systems, and energy capacity. Kharbanda stressed that enterprise AI adoption will attract substantial spending as companies move from pilot projects to production deployments across sectors such as banking, telecom, manufacturing, healthcare and retail. According to them, sector-specific AI applications will continue to grow because that is where business value is ultimately realised.
