- Indian IT faces client budget caution impacting transformation deals.
- US tech booms on AI infrastructure, not Indian IT services.
- Indian IT participates in AI, lacking model/infrastructure ownership.
The Nifty IT index is down 23.34 per cent year-to-date and 19.14 per cent over the past one year. In the same period, US technology stocks are experiencing one of their strongest runs in recent memory — Google committing USD 40 billion into Anthropic, Microsoft reporting AI revenue of USD 37 billion growing at 123 per cent year-on-year, Nvidia printing record revenues quarter after quarter. The contrast is jarring enough that serious investors are asking a question that would have seemed absurd three years ago: is Indian IT structurally finished?
The honest answer requires separating three distinct questions that are being conflated in the current panic what is happening right now, what is structurally different about Indian IT versus US AI companies and what the medium-term trajectory actually looks like.
Why Indian IT Is Suffering
When US Tech Is Booming The immediate cause of Indian IT underperformance is client-side budget caution in the United States. Indian IT companies derive the bulk of their revenues in some cases 80 per cent or more from US enterprise clients. When US corporates face uncertainty from tariffs, interest rate ambiguity or macro concerns, discretionary technology spending is the first budget line to get deferred.
The deals that get pushed out are exactly the kinds of large transformation engagements that drive Indian IT revenue growth. Meanwhile, the US technology companies that are booming are not spending on Indian IT services. They are spending on their own infrastructure — data centres, GPU clusters, model training, chip design. This capex goes to Nvidia, Construction companies, power utilities and to each other. The Magnificent Seven are investing in each other’s AI ventures in an increasingly interconnected investment ecosystem.
Google invests in Anthropic. Microsoft owns 27 per cent of OpenAI. Amazon backs Anthropic alongside Google. These are not customers of Indian IT they are competitors to each other in infrastructure arms race that Indian IT companies are not directly participating in at the core infrastructure and model development phase. The boom in US tech is real but narrow. It is concentrated in AI infrastructure spending among a handful of companies who are simultaneously investors, customers and competitors in the same ecosystem. The broader US enterprise technology budget the one that actually flows to Indian IT has not boomed in the same way.
Also Read : UPI Cashback Isn’t Always ‘Free Money’: Here’s What Users Miss
The AI Circle and Whether India Is Right to Stay Out
Capex announcements from big tech also serve as strategic signals that influence competitive behaviour. When Google announces USD 50 billion in AI capex, OpenAI and Anthropic have to raise even more capital just to remain competitive. The labs are on a permanent fundraising treadmill where each round needs to be larger than the last and the pool of investors who can write those cheques is getting smaller every quarter. OpenAI is projecting around USD 25 billion in annual revenue alongside losses estimated in the double-digit billions.
That is not a company struggling with product market fit. That is a company where the fundamental cost of running frontier models consumes more than half the revenue it generates. Energy at multi-year highs, Gulf capital pulling back. OpenAI’s recent decision to run advertisements inside ChatGPT, which Sam Altman once described as a last resort, tells you something about where the financial pressure sits. Meta raised its full year 2026 AI capex guidance to between USD 125 billion and USD 145 billion when it reported Q1 results up from the already ambitious USD 115 to USD 135 billion range it had set at the start of the year, driven by higher component costs and an accelerating push toward what the company is now openly calling superintelligence infrastructure.
To put the scale in perspective, Meta spent USD 72 billion on capex in all of 2025 the new target nearly doubles that in a single year. Google’s cloud revenue grew 63 per cent to USD 20 billion in a single quarter and its contracted future business backlog doubled to USD 462 billion. Microsoft’s AI business is at a USD 37 billion annual run rate growing 123 per cent year-on-year. If this funding cycle were to reverse and like every previous technology financing cycle has broken before the technology itself failed the impact could be significant.
Banks that financed data centre construction on growth projections that may not materialise, pension funds holding the Magnificent Seven at peak valuations, GPU manufacturers sitting on inventory ordered for demand that did not arrive. The technology survives. The question is who survives with it. Indian IT’s conservative capital allocation — paying Dividends and doing buybacks rather than betting billions on frontier model development looks like timidity from the outside right now. It may look like wisdom in hindsight.
Where Indian IT Actually Sits in the AI Value Chain
This is the section the current narrative around Indian IT gets most wrong and it matters enormously for how investors should think about the sector. Indian IT is not absent from AI. It is present at every layer except the most capital-intensive and loss making one. The framework is cleaner than most coverage suggests. Layer Who Dominates India’s Position Frontier Models (LLMs) OpenAI, Google DeepMind, Anthropic Not present Infrastructure (GPUs, Data Centres) Nvidia, Hyperscalers Limited Implementation and Services IT Services companies Dominant Indian IT companies are already generating real AI revenue. TCS reported annualised AI services revenue exceeding USD 2.3 billion in Q4 FY26, driven by mega deals and enterprise adoption, with a total contract value of USD 12 billion. HCLTech’s advanced AI revenue annualised at USD 620 million in Q4 FY26, implying a quarterly run‑rate of about USD 155 million, up roughly 6 per cent quarter‑on‑quarter, with AI‑related work now forming a significant part of new large deals.
Infosys closed large deals with a total contract value of USD 4.8 billion, with its Topaz AI platform acting as a key differentiator, and the company has raised or reiterated its full‑year guidance amid AI‑driven momentum. Wipro is pivoting to AI‑native and AI‑infused solutions, with over 25,000 employees being upskilled in Microsoft Cloud and related technologies, including Azure and AI‑powered tools. The partnerships underpinning these numbers are with the very companies dominating AI globally.
TCS has deepened its Google Cloud collaboration and deployed ChatGPT Enterprise through a partnership with OpenAI. Infosys has a strategic collaboration with Anthropic for enterprise AI agents across telecom, finance and manufacturing, and a separate partnership with OpenAI for AI-driven software development. HCLTech is building custom AI agents using Google’s wGemini with 23,000 professionals being trained. Wipro has committed to delivering 200 production-ready AI agents across healthcare, banking and retail through its Google Cloud partnership. Indian IT is not waiting for AI to arrive. It is building on top of the models that US and Chinese companies are building, deploying those models inside enterprise clients and generating contracted, cash-generative revenue from that deployment. What it does not do — and this is the real gap — is own the models, the compute infrastructure or the product-led monetisation that comes with them.
The Three Missing Layers
The gap in Indian IT is not just the absence of a large language model. It is three things simultaneously. The first is frontier model ownership. There is no Indian equivalent of GPT-4 or Claude. India generates approximately 20 per cent of the world’s data but foreign models are being trained on it. R&D spending in India is below 1 per cent of GDP versus South Korea at 5 per cent and China at approximately 2.5 per cent. A country that does not invest in foundational research does not build foundational models. The second is infrastructure ownership.
Indian IT companies do not own GPU clusters, hyperscale data centres or the training compute that determines who can build and run frontier models. The data centre buildout underway — Reliance’s Jamnagar campus, AdaniConneX, Airtel’s Nxtra will address the compute gap over five to seven years. The research talent gap is a harder problem — India’s best AI researchers have largely been absorbed by US companies through compensation that Indian organisations cannot match. The third is product-led monetisation.
US AI companies sell AI as platforms and products with scalable margins. Indian IT sells services around AI with linear growth economics. Product revenue scales without proportional headcount growth. Services revenue largely does not. This structural difference is what a senior Microsoft executive captured precisely: the era of the headcount driven model is over. Adding half a million people no longer doubles revenue. Whether Indian IT can increase profitability per person at the pace previously driven by headcount growth is the central question the sector has not yet answered.
Also Read : India Housing Sales Dip 7% As West Asia Conflict Weighs On NRI Sentiment
Is Indian IT Different from US AI Companies
Fundamentally yes, and the difference matters more than the current price action suggests. US AI companies are making speculative bets on technology that does not yet have proven unit economics at scale. They are spending capital today in the hope that commercial AI applications create revenue streams large enough to justify the investment. Some will be right. Some will not. The ones that are wrong will experience significant value destruction — analysts are already flagging Meta’s projected USD 24 billion negative free cash flow by 2027. Indian IT companies deploy proven technology for paying clients under contracts with defined deliverables. TCS does not lose money on a deal. Infosys does not build a product and hope someone buys it.
The revenue model is services-based, contract-driven and cash-generative. This produces lower peak valuations in bull markets and significantly lower downside in corrections. The USD 2.3 billion AI revenue TCS is generating today is real, contracted and growing not a projection based on a hoped-for future market. When Does Indian IT Come Back The cloud computing parallel is worth taking seriously. When US companies began their massive cloud infrastructure buildout in 2015 and 2016, the Nifty IT index went sideways and negative through 2016 to 2018. It looked structurally impaired. Then Indian IT started winning cloud implementation, migration and management contracts and from 2018 to 2020 the index delivered one of its best runs.
The same sequencing is playing out with AI. The first phase — infrastructure buildout, model development, foundational research is being funded and executed by US and Chinese companies. The second phase enterprise AI implementation, model integration, workflow transformation, security and compliance is where Indian IT has always competed and where its current AI revenue is already being generated. Every large enterprise that adopts AI needs help deploying it safely, connecting it to existing systems and managing the regulatory implications.
That work is not done by OpenAI or Google. It is done by the implementation partners. Indian IT firms are already those partners the question is whether the implementation market scales fast enough and at high enough margins to compensate for the headcount growth engine that is structurally diminished. Indian IT is not missing AI. It is participating in AI without owning the most valuable parts of it. That is the honest picture. The story is not over the current phase simply belongs to a different set of players. The next one is where Indian IT has always built its strongest returns.
(Disclaimer: This article uses information originally published by Dalal Street Investment Journal (DSIJ). The views expressed are those of the original authors and not necessarily of ABP Network Pvt. Ltd. This content is provided for general informational and educational purposes only and should not be construed as investment, financial, legal or tax advice. Readers are advised to conduct their own research and/or consult a qualified financial advisor before making any investment decisions. This content is for informational purposes only and should not be treated as investment advice. ABP Network, its employees and associates shall not be responsible or liable for any losses or damages arising directly or indirectly from the use of or reliance on this article or any information contained herein.)
