India’s IT bellwether Tata Consultancy Services has kicked off one of its most ambitious bets yet — a $6 billion foray into artificial intelligence infrastructure that marks a decisive pivot from its traditional services-led model toward the capital-heavy world of AI data centres. Just when India’s IT services giants were being blamed widely for having been left out in the AI boom, TCS declared ambitions to become the world’s largest AI-led technology services company.
The company plans to build a sovereign, co-location AI data centre of up to 1 GW capacity in India over the next five to seven years through a new subsidiary. But with returns uncertain and synergies with its core services unclear, the $6 billion question is whether this bold leap will redefine TCS’s growth story or test its financial discipline.
Analysts remain split. Some call it a strategic move to future-proof the business as global AI demand explodes; others warn it’s a low-margin, high-capex detour that could dilute TCS’s stellar return profile.
The initiative signals a rare shift in strategy for the usually conservative IT giant, putting its balance sheet to work at a time when the industry is chasing AI scale. TCS will follow a co-location model, where it will provide the passive infrastructure and clients will bring in compute and storage. The company expects the capital intensity to be approximately $1 billion per 150 MW, with funding to be structured through a mix of equity and debt, supported by financial partners, TCS said in an analyst call after declaring its Q2 numbers on Thursday.
Management said the first phase will become operational in 18–24 months, with initial anchor clients from hyperscalers, deep-tech AI firms, Indian enterprises, and sovereign projects. TCS noted that India’s current installed data-centre capacity stands at approximately 1.2 GW, but demand could expand nearly 10x in the next five to six years, while committed capacity is only 5–6 GW, creating a significant revenue opportunity.
Is the AI boom becoming a bubble? Why Goldman, JPMorgan, IMF are sounding the alarm
Dalal Street isn’t entirely convinced. Akshat Agarwal of Jefferies said the move was surprising for three reasons. First, it doesn’t materially change TCS’s growth profile since at full potential in FY31, revenues from the business will be approximately $1.2 billion and add only 0.6% revenue CAGR over FY25-31. Second, it’s a physical infrastructure leasing business with little relationship to TCS’s existing IT service business. Third, it’s a capex-heavy, low ROCE (8–12%) business, quite the opposite of TCS’s current business.
Jefferies maintained its Hold rating with a price target of Rs 3,100 based on 21x PE, cutting FY26–28 EPS estimates by up to 1% and expecting only a 4% EPS CAGR over FY26–28E. The firm noted it has not modeled the impact of the data centre foray in its numbers, though the likely impact on free cash flow and payouts may be less than 10% due to the use of debt funding.
Motilal Oswal highlighted that TCS will not run cloud workloads or provide managed cloud services — the data centre will function as a sovereign co-location site. “This means low technology intensity, limited overlap with TCS’s core services portfolio, and hence minimal direct synergies,” the brokerage noted.
The firm added that ROE at the subsidiary level will be lower than TCS’s over 50% group ROE, given the capital-intensive nature. However, management guided that at the consolidated level, it will not be margin-dilutive, as external partners will share funding. Motilal Oswal said the move is “best viewed as cash deployment rather than a services-led growth driver.”
TCS Q2 Results: Cons PAT rises 1.4% YoY to Rs 12,075 crore, misses street estimates
Nomura questioned the return profile, noting it is “unclear to us how the ROEs of this business will be similar to existing business RoE of ~50% given its capex-heavy nature with IRRs at ~20%.”
However, not everyone is bearish. Nuvama called the decision “a balanced one, as it explores growth opportunities in the tech ecosystem, with manageable capital commitment (for TCS’s balance sheet size) and decent returns profile (though lower than its own).” The brokerage acknowledged TCS is “deploying significant capital, in high-growth but lower RoCE business, than its own,” but views it as an acceptable trade-off.
JM Financial struck the most bullish tone, calling the co-location foray “value accretive,” though it sees “limited synergies with its Services business at this stage.” Importantly, the brokerage said, “TCS’ willingness to put its balance sheet to use, a significant deviation from its earlier conservative approach, is a welcome change in the current dynamic environment. Investors should encourage this.” JM Financial maintained its Buy rating.
The Five Pillars of AI Strategy
Beyond the data centre, management laid out five pillars of its AI strategy: making TCS AI-first by enabling employees to learn, experiment, and embed AI in daily work; redefining every service line under a “human + AI” delivery blueprint; building a future-ready talent model by investing in future-ready skills and recruiting top talent locally; making AI real for clients through rapid builds, AI labs and offices, and value-chain solutions across industries; and strengthening ecosystem partnerships while stepping up efforts in M&A and scaling AI platforms — including the recent acquisition of ListEngage.
The company is targeting an annuity stream of revenue from pure-play AI players, deep-tech, hyperscalers, and private and government enterprises in India.
While TCS posted a decent quarter on low expectations, its decision to invest in an AI data centre takes it to an interesting crossroad — one where its legendary financial discipline meets the capital-hungry demands of the AI era. Whether this gamble pays off may define the company’s next chapter.
as a Reliable and Trusted News Source
Add Now!
