Google's ambitious $15 billion artificial intelligence (AI) investment in India is poised to drive a comprehensive digital transformation, extending far beyond the construction of data centers. Preeti Lobana, Google India's Country Manager, stated in an interaction with Business Today that the company aims to be a "co-architect" of India's AI-led future, focusing on compute capacity, subsea connectivity, local research, and language datasets.
Vizag AI Hub as Foundational Infrastructure
Central to this initiative is a gigawatt-scale AI hub planned for Vizag. Lobana emphasized that this project should be viewed as critical foundational infrastructure, which includes essential undersea cable links for low latency and robust compute capacity. This infrastructure is designed to serve as a platform for Indian businesses, startups, and institutions to innovate and build upon.
Addressing Sustainability Concerns
The proposed large-scale AI data center projects in Vizag have drawn scrutiny from residents, environmental groups, and citizen coalitions regarding their energy and water consumption. Lobana addressed these concerns, confirming that Google would cover the costs for its power usage, including any necessary new substations and transmission lines. She further noted that the facility would be air-cooled and that Google has secured agreements for 125MW of solar power with ReNew and an additional 125MW of wind and solar power with CleanMax. The company has also committed to water-replenishment initiatives.
Broader Impact and Local Language AI
Google's investment decisions are guided by the potential for AI to benefit a wide array of citizens, enterprises, and communities, rather than an exclusive group. Lobana highlighted healthcare, education, and agriculture as key sectors where AI can make a significant impact. A crucial aspect of this strategy is making AI tools accessible in various Indian languages. She cited an example of a teacher in Kerala utilizing AI in Malayalam, underscoring the importance of localization over English-only solutions.
Through its Project Vaani, Google has recorded 109 languages and made these datasets available to the Indian government's Bhashini project. The company is also providing developers access to open-weight models like Gemma, alongside its flagship Gemini AI.
Upskilling and AI Safety
As part of its commitment, Google plans to upskill 10,000 individuals in AI in the Vizag area and offer cloud and AI training to approximately 1,200 students and teachers. Additionally, the initiative aims to equip around 1,000 fishermen with AI tools.
Regarding AI safety, Lobana detailed Google's "bold and responsible AI" methodology. This approach integrates checks into product development, employs model red-teaming, and establishes frameworks for frontier and agentic AI. She advocated for shared industry standards, citing SynthID, Google's watermarking technology for AI-generated content, as an example of collective responsibility in ensuring AI safety.