Sitharaman Pitches India as an Alternative to the US and China for AI and Chip Capital

India wants to become a third pole for global investment in AI and hardware. Finance Minister Nirmala Sitharaman said the country is a viable alternative to the US and China for international capital, and named three strengths: artificial intelligence, semiconductors and green technology. The claim rests on a chip-support program of roughly $10 billion and the state IndiaAI mission, which has a budget of about $1.25 billion. The Bloomberg report of 13 October 2026 is short and gives no details of the speech, so below we lay out the claim and the context it relies on.

What exactly was said

The wording is cautious: "alternative," not "replacement." Sitharaman does not claim that India will displace American or Chinese markets. She talks about diversification. Funds that have already invested in the US and carry China risk get a third option, with a huge domestic market and a growing technology base.

Notably, AI appears in the list next to chips and energy. For an investor, this is a linked chain: models need data centers, data centers need electricity and server hardware, and hardware needs manufacturing. India is trying to show that it covers all three links, though to varying degrees of readiness.

What backs the claim

The publicly known facts look like this (amounts are as stated by the government and companies):

• India Semiconductor Mission: subsidies of roughly $10 billion, approved in December 2021.
• IndiaAI Mission: a budget of 10 372 crore rupees (about $1.25 billion), approved in March 2024, which includes state GPU computing capacity for startups and researchers.
• Micron: an assembly and test facility in the state of Gujarat, with an investment of around $2.75 billion.
• Tata Electronics with PSMC: a fab in Dholera valued at about $11 billion.
• Google: a stated $15 billion over five years for an AI hub in Visakhapatnam.
• Microsoft: a promised $17.5 billion in investment in India through 2029.

This makes a good set of slides. But announced sums and working capacity are different things.

Where the claim runs into reality

The first constraint is energy. Data centers for training and inference consume megawatts, and reliable power and cooling in a hot climate are expensive. Grids in a number of regions are under strain, so the speed at which new campuses get connected will be a critical metric.

The second is timelines in semiconductors. The first Indian projects focus on packaging and mature process nodes, not leading-edge ones. That is a normal start, but India is far from competing with Taiwan and South Korea on top-end chips. A fab needs years, materials suppliers and engineers who still have to be trained.

The third is conditions for capital: taxes, data regulation and predictable rules. Here India will have to compete not only with the US and China but also with Vietnam, Malaysia, the UAE and Saudi Arabia, which also offer subsidies and fast permitting.

What to watch next

If you follow the industry or decide where to place capacity, look at verifiable signals rather than rhetoric:

• The launch of the first serial chips from Indian lines and their actual volumes.
• How quickly new data centers get connected to the grid, and what energy tariffs look like.
• The share of announced investment that has turned into built sites.
• The emergence of local models and startups that use the state IndiaAI GPU cluster.
• Changes in tax and regulatory conditions for foreign investors.

For developers and AI companies, the practical takeaway is simple: India already matters as a market and a source of engineers, and it will become significant as a compute location as power and hardware come online. Watch the commissioning figures, not the volume of promises.