Compute capacity is foundational infra of the AI era, says Preeti Lobana, vice-president and country manager at Google India

Compute capacity is the foundational infrastructure of the modern era - akin to national power grids and railroads that fuelled gross domestic product during the industrial revolution, said Preeti Lobana, vice-president and country manager at Google India during a fireside chat with ET's Surabhi Agarwal.

Discussing Google's major commitments in India, including a $15 billion investment toward building a 1-gigawatt AI data centre hub in Visakhapatnam, Lobana outlined how a combination of cutting-edge hardware, localised AI models, and inclusive upskilling would power India's ambitious target of a $1 trillion digital economy.

"When we talk about industrialisation, countries invested in building railroads and highways and power grids and that is what powered GDP in that time," Lobana said. "When we talk about the AI era, data centres, the compute infrastructure is what the new national grid is to power our economy. For us, when you look at the India AI Mission as well, the government has clearly stated that compute capacity is very critical and foundational to sort of drive this momentum forward in the AI era."


Indian organisations have swiftly moved past early proof-of-concept stages directly into full-scale operational deployments, she said. Citing Tata Steel as an example, she explained how the steelmaker deployed 300 specialised AI agents within nine months to optimise invoicing and elevate shop-floor safety.

Also Read | AI, trust, and the future of money: With AI making payments invisible, trust becomes the new currency

Speaking about why she is bullish on the "thriving" Indian startup ecosystem, Lobana said many founders in tier-2 and tier-3 cities were also solving for and finding users across the globe.

"India has the highest number of funded consumer AI companies," she said. "That tells you the scale of ambition that the founders have."

Foreign-developed AI models cannot simply be imported without deep local adaptation, Lobana said. Through initiatives like Gemini Live, which is now supporting 25 Indian languages, and partnerships with institutions like the Indian Institute of Science (Project Vaani) and the government's Bhashini initiative, she said Google is digitising dialectical nuances in 773 districts to make AI accessible to non-English speakers and low-resource language communities. This is critical not just for commercial success, but also for using AI for social good, she said.

Lobana framed AI as a force multiplier and called for public-private partnerships to upskill workers across sectors. She said the opportunity lies in empowering those in tier-2 and tier-3 cities, like small enterprises, farmers and workers across the "heartbeat of India".

"It is about using AI to empower all of those people so that the economic value and benefits of AI empowers them...AI for all, not just a chosen few," she said.

"You can use AI to upskill people much faster and scale better than you could pre-AI, but it's also about how to use AI to be far more productive and effective," Lobana said. "It is about making sure that we are all not using AI just as a chatbot or as a productivity tool, but truly how can you learn, be very comfortable so that you stand out from others, because it's about how you use it to maximise your output and your creativity."