India's AI market to boom to USD 131 billion by 2032 at 39% CAGR; Deployment of up to 700,000 GPUs by 2031: Report

According to a report by Avendus Capital, within the 2025 market segment split, the BFSI sector leads with USD 2.5 billion, followed by startups at USD 1.8 billion, media at USD 1.6 billion, manufacturing at USD 1.4 billion, and tech services at USD 1.3 billion. The public sector accounts for USD 1.2 billion, while other segments make up USD 3.3 billion.

As per the report, India's AI data centres could see deployment of up to 700,000 GPUs over the next 5 years, unlocking a USD 23 billion investment opportunity. This rapid expansion stems from a compounding growth rate in the domestic artificial intelligence industry, which projects to expand at a 39 per cent compound annual growth rate (CAGR) over the next seven years.

Enterprise infrastructure choices support this shift, as 64 per cent of enterprises indicate the propensity to build in-house AI products on cloud-provisioned compute and storage resources. The country currently hosts more than 1,800 Global Capability Centres (GCCs), which includes over 500 centres focused entirely on AI. Furthermore, India acts as a base for approximately 1.8 lakh startups, where roughly 89 per cent of newly launched startups utilize AI within their services.

As per the report, India maintains the second highest number of ChatGPT users globally with a 9 per cent share, trailing only the United States which holds 18 per cent. Brazil, Germany, and the United Kingdom follow with 5 per cent, 4 per cent, and 3 per cent respectively, while other nations account for the remaining 61 per cent.

This high consumer adoption aligns with enterprise readiness, where 45 per cent of Indian enterprises already implement AI in their business processes. Within this group, 25 per cent represent small-scale implementation intended to scale over the next two years, and 20 per cent operate with AI-led processes aligned with business metrics.

Meanwhile, 49 per cent of companies remain in the proof-of-concept phase, and 6 per cent state that AI is not a priority or remain yet to explore its applicability.

The IndiaAI Compute Pillar aims at providing high-end GPUs at affordable prices. This pillar has received a total commitment of over 38,000 GPUs from bidders, with approximately 58 per cent of the allocation, or around 22,000 GPUs, already allocated for AI workloads.

Additionally, the AlKosh dataset platform focuses on developing large datasets for training indigenous AI models, allowing developers to focus directly on solutions rather than basic modules. The platform features 3,000 datasets and 243 AI models across 20 distinct sectors.

Under the IndiaAI Foundational Models initiative, which focuses on developing localized Large Language Models (LLMs) using local data and languages, the mission received over 500 proposals from model developers. Selected developers for these initiatives include Sarvam, Gnani.ai, Soket, and Gan.AI. (ANI)