Opinion Google’s $15 billion AI investment in Andhra Pradesh: A boost, a beginning

The last few years have seen a dramatic surge in AI investments, driven by global tech majors such as Microsoft, Alphabet, Meta, Nvidia and OpenAI. As developing AI models and then making them available for use at a global level requires huge capital, the scale of the investments being made has reached impressive proportions. In 2024, global corporate investment in AI touched $252.3 billion as per the AI index report from Stanford University. As per McKinsey, the data centres needed to handle AI processing are expected to require $5.2 trillion in capex by 2030, the bulk of which is likely to be directed towards technology developers to produce chips and computing hardware. While some have expressed concerns over this AI boom — the IMF has, for instance, warned that “the risk is also that lofty profit expectations will ultimately be unmet” — others are more optimistic with tech majors stepping up their investments.

Earlier this week, Google announced that it will invest $15 billion over the next five years (2026-2030) to set up an AI data centre in Andhra Pradesh. Situated in Visakhapatnam, it will be Google’s largest outside the US, and will include a “purpose-built data centre campus capable of gigawatt-scale compute capacity”. Earlier this year, during its second quarter earnings season, the company had raised its capex forecast for the year to $85 billion to meet the growing demand as adoption rates for both firms and companies surge. This is a big boost to India’s plans to be a major player in this space. India’s data centre industry had crossed the 1 GW mark in 2024 as per a report by JLL. Estimated to account for 52 per cent of the country’s capacity, Mumbai leads the market, followed by Chennai and Delhi.

Advertisement

The world over governments are announcing ambitious plans and pledging huge resources in the AI space. The Indian government has also taken several steps to help build the larger ecosystem in the country. Ten semiconductor manufacturing projects, including high-volume fabs and OSATs (outsourced semiconductor assembly and test), have been approved. Alongside, 23 chip-design projects have been sanctioned by the government and 12 startups have been chosen to build indigenous large language models and small language models. Thus, across the spectrum, from the physical infrastructure (data centres) and the semiconductor projects (chips) to developing large language models, India is making progress. These efforts need to be supplemented by a similar focus on boosting research and development and skilling the labour force.