Nirmal Bang also said employee additions at large-cap IT companies remained broadly flat, with total headcount rising just 0.1 per cent on average. It said this indicates relatively low exposure to AI-led services delivery. Mid-cap IT services companies, however, performed better than their large-cap peers, broadly in line with the brokerage’s expectations. It said smaller market share leaves them with less legacy business to defend, while some have won deals previously beyond their reach by using AI-led delivery models.
The report said the gains from generative AI are increasingly flowing to AI model companies rather than IT services providers. As AI improves developer productivity, work that earlier generated billable engineering hours can be completed with fewer human hours, even as computing and token consumption rises. Nirmal Bang cited OpenAI’s quarterly revenue growth of 130 per cent year-on-year to US$6.7 billion and Anthropic’s revenue of US$11.5 billion, up 1,300 per cent year-on-year, as examples of the rapid expansion of the AI model layer.
Nirmal Bang estimates developer throughput could rise 5-8 times, while clients may build around 20 per cent pricing compression into technology budgets. It said this could shift part of the economic value historically captured by IT services companies towards compute and AI model providers. The brokerage expects this pressure to become more visible across the sector as AI-led productivity gains increasingly affect traditional services revenue.
The brokerage sees two possible sources of stress for IT services in 2026. In one scenario, rapid AI adoption could reduce demand for traditional effort-based services as generative AI automates software development and testing, lowering incremental manpower needs and increasing pressure on pricing and outcome-based contracts. In the other, if AI adoption slows or does not deliver measurable returns, global technology budgets could come under pressure, especially given the large capital commitments being made by hyperscalers to AI infrastructure.
Nirmal Bang expects large-cap IT companies to face greater pressure in FY27 because of their bigger legacy revenue bases and higher exposure to potential wallet-share leakage towards AI model companies. It sees relatively lower risk for mid-cap firms such as Coforge and Persistent because of their smaller scale, greater agility and lower exposure to renewal-led compression affecting larger players.
It also cautioned that rupee depreciation may provide an optical boost to FY27 earnings, but this may not translate into strong EPS growth if delivery economics and EBIT margins remain under structural pressure. Nirmal Bang maintained an underweight stance on the IT services sector, saying the long-term opportunity lies in a leaner, AI-native services model, but the transition for traditional players is likely to be painful and front-loaded.
Nirmal Bang has 'sell' ratings on TCS (Target Price: Rs 1,724), Infosys (Target Price: Rs 1984), Wipro (Target Price: Rs 151), HCL Tech (Target Price: Rs 949), Tech Mahindra (Target Price: Rs 1,187) and LTM (Target Price: Rs 3,235), and 'hold' ratings on Coforge (Target Price: Rs 1,971), Persistent Systems (Target Price: Rs 5,734) and Mphasis (Target Price: Rs 2,375).