AI transforms Indian IT contracts as clients seek greater value

Artificial intelligence is rapidly transforming India’s $315 billion IT industry, prompting outsourcing giants such as Tata Consultancy Services (TCS), Infosys, Wipro, HCLTech, and Cognizant to overhaul their business models. As global clients push aggressively for greater productivity and cost efficiencies, IT service providers are increasingly shifting from traditional billable hours toward fee structures tied to clear, measurable performance outcomes.

Where once a large employee base provided a decisive advantage for winning deals, automation enabled by AI is reducing the need for armies of entry-level coders, narrowing the gap between large and mid-sized firms. Industry insiders report that clients are now insisting on steeper price cuts and expecting contracts to guarantee productivity gains, driven by AI’s promise of efficiency. This has resulted in shorter contract durations, contracts lost to clients implementing AI in-house, and heightened competition from more agile, smaller rivals able to deploy senior staff and offer flexible terms.

Jimit Arora, CEO of the consultancy Everest Group, described market conditions as “desperate” for service providers, noting that the negotiating power has decisively shifted in favor of clients. The Nifty IT index has dropped by 20% so far this year, with its 10 listed companies shedding a combined $73 billion in market value amid mounting concerns that AI is eroding their core businesses.

Within TCS, Chief Executive K Krithivasan told Reuters that around 80% of contracts in finance, HR, and business services segments are now based on outcome performance measures. According to a source, this is double the proportion seen before AI went mainstream in late 2023. Various companies are reporting similar approaches: an AI and automation deal between Cognizant and Daimler Truck in February, for example, stipulates that AI-driven cost savings will be split between the two companies. Cognizant acknowledged a shift toward value-focused and outcome-based contracts, as clients demand more measurable returns for their spending.

Meanwhile, HCLTech’s multiyear agreement signed in June 2025 with Germany’s E.ON will see the IT company unpaid for the first year, with future payments linked to realized efficiency gains and specific outcomes. Persistent Systems CEO Sandeep Kalra said clients now expect the same work for 25% to 30% less, alongside faster delivery and higher productivity. On the upside, AI is letting Persistent compete for and win larger deals than before, and helped drive a 16% rise in revenue last quarter. Rival Coforge’s sales jumped by a third in the same period. For contrast, the largest IT providers saw much slower revenue growth of just 1% to 3%.

Analysts and executives alike point to intensifying competition from mid-sized firms that can respond quickly and price flexibly. “Many Tier 2 firms have been more agile and hungry in this phase,” said Phil Fersht of HFS Research.

However, some leaders warn of overreach in the scramble for market share. Tech Mahindra CEO Mohit Joshi observed that some competitors are making risky promises by projecting 70% to 80% productivity gains over five to seven years and guaranteeing prices despite higher chip costs. Infosys has walked away from deals it considered uneconomic. For now, TCS says it is compensating for revenue pressure from AI with new business but acknowledges that future growth depends on staying ahead of the “revenue deflation.”

Reflecting the changing landscape, last year TCS implemented layoffs of more than 12,000 employees, the first major Indian IT company to do so in the AI era, and industry leaders now acknowledge that the traditional mass recruitment of entry-level engineers will likely no longer be needed. “The pyramid model is gone. With coding agents, we no longer need basic coding,” said former Infosys CFO V. Balakrishnan.

As AI continues to reshape how IT services are delivered, India’s outsourcing industry faces a defining period of reinvention.