Give AI investment the big come hither

AI Impact Summit starts in New Delhi today at a time when AI investments globally are now in uncharted territory, with the US, China, the EU and Asia-Pacific entering the hyperscaler range. The five biggest AI infra providers - Amazon, Alphabet, Microsoft, Meta and Oracle - are committing nearly $700 bn in investments in 2026, twice the amount they spent in 2025. Apart from this, the US's Project Stargate has a $500 bn strategic investment plan involving OpenAI, SoftBank and Oracle. In China, Alibaba is committed to investing $53 bn over a 3-year period, ByteDance has $13 bn earmarked for AI infra in 2026 while Tencent has trimmed its investments.

China's AI investments, although not on the US scale, are climbing as it pursues a less capital-intensive development model. The EU has unveiled a ₹240 bn (about $260 bn) action plan for AI infrastructure. Japan has allocated 1 tn yen (about $6.5 bn) annually for AI and semiconductor development and South Korea has lined up 9.9 tn won ($6.8 bn) in its national AI budget. Besides these, a clutch of sovereign wealth funds in West Asia are investing huge amounts to kit out local AI infra. Economies that have hyperscaled their AI investments are reporting that capacity is being absorbed quickly. Energy requirements for AI data centres are emerging as a key constraint. This makes the case for investment diversification.

India has tweaked its tax regime to plug into the AI investment tsunami. The budget announced a 20-yr tax holiday for foreign cloud service providers operating through India-based data centre infra. Investments of $70 bn in Indian data centres are in various stages of implementation, and $90 bn more has been announced. The AI policy is meshed with other GoI initiatives to push semiconductor and electronics manufacturing. The country is positioning itself as a long-term destination for AI investments, from all sources. This holds the key to its emergence as an AI power.