How artificial intelligence is dismantling the job ladder that built India’s middle class, and forcing the world’s IT industry to confront a question it never expected to ask itself
For three decades, the unspoken promise of India’s technology sector was this: learn to code, get placed, and you will never have to worry about money again. Tata Consultancy Services alone grew from roughly one lakh employees in 2007 to over six lakh by 2023, its headcount swelling almost in lockstep with India’s aspirational middle class. That promise has now been withdrawn quietly, and without apology.
In July 2025, TCS announced it would cut 12,261 employees, 2% of its global workforce, in what the company diplomatically called a ‘skill realignment.’ Its last comparable cut, in FY15, affected only 1% of staff. CEO K. Krithivasan told Moneycontrol that the decision was driven by skill mismatches, not by AI productivity gains. That careful clarification has not reassured the industry. When India’s largest private sector employer lays off 12,000 people, the reason it gives matters far less than the direction it signals.
The direction is unmistakable. Across TCS, Infosys, Wipro, HCLTech, Tech Mahindra, Cognizant, and Accenture’s India operations, more than 25,000 tech jobs were cut in 2025. This is not the bottom rung being sawn off. This is the middle of the ladder.
The Pyramid Was Always Fragile
The labour model that made Indian IT dominant was elegantly simple: hire thousands of freshers cheaply, bill them to Western clients by the hour, train them over years, and gradually move them up a pyramid. It worked because people were cheaper than software. That arithmetic has inverted.
Infosys CEO Salil Parekh disclosed in earnings calls that the company has deployed over 300 enterprise-grade AI agents delivering 5% to 15% productivity gains, enabling revenue growth without proportionate headcount increases. Wipro has deployed more than 200 AI-powered ‘intelligent agents’ in HR, finance, and legal. Chairman Rishad Premji describes generative AI as a ‘game-changer’ and says the company is moving toward fully autonomous systems in some functional areas. The message from clients reinforces this: a Jefferies analyst report noted that customers are now structurally asking vendors to deliver ‘the same work with fewer employees.’ When that becomes a contractual expectation, headcount reductions are not a choice, they are arithmetic.
The World Economic Forum’s Future of Jobs Report 2025 predicts AI will eliminate 8.3 core jobs globally by 2027 while creating 6.9 crore, a net loss of 1.4 crore. In the US, consulting firm Challenger, Gray & Christmas attributed nearly 55,000 layoffs directly to AI in 2025, out of a total 11.7 lakh cuts, the highest level since the pandemic year of 2020. Over 1,84,000 global tech jobs were cut by October 2025, of which more than 50,000 were explicitly linked to AI and automation. Amazon eliminated 14,000 corporate roles in October, citing AI-enabled leaner structures. Microsoft cut approximately 15,000 jobs through the year. Salesforce reduced its customer support workforce from 9,000 to 5,000 using agentic AI, a fact CEO Marc Benioff stated without evident discomfort.
The Entry Point Has Closed
The cruelest irony of this restructuring is not falling on the experienced professional alone. The Burning Glass Institute tracked a sharp contraction in entry level hiring across AI-exposed fields between 2018 and 2024: the share of software development jobs requiring three years of experience or fewer dropped from 43% to 28%. For data analysis, the figure fell from 35% to 22%. Total job postings in these fields stayed flat or rose. Companies are not hiring fewer people; they are skipping a generation entirely.
Research from SignalFire shows Big Tech companies reduced new graduate hiring by 25% in 2024 compared to 2023. January 2025 saw the lowest job openings in professional services since 2013, a 20% year-on-year drop. 40% of white collar job seekers in 2024 failed to secure a single interview, according to analysis by MyPerfectResume. Anthropic CEO Dario Amodei has said AI could eliminate half of all entry level white collar jobs within five years. Nvidia CEO Jensen Huang pushed back at VivaTech 2025, arguing that greater productivity historically leads to more hiring, not less. Professor Dilan Eren of Ivey Business School offered the sharpest structural critique: eliminating junior roles to cut costs is ‘an exponentially bad move’ that destroys the internal talent pipeline firms rely on for their next generation of senior engineers.
India feels this acutely. The country produces over 15 lakh engineering graduates annually, according to multiple industry estimates. The IT sector, which contributed approximately 7.5% to GDP in FY23 and employed 56.7 lakh people as of March 2025, has been the primary absorber of this graduate output for 30 years. Sonal Varma, chief economist for India and Asia ex-Japan at Nomura, stated plainly: “Entry-level routine jobs are being displaced, and mid-level jobs are transforming.” That transformation is happening faster than any reskilling programme can track.
Reskilling’s Hollow Promise
Every firm caught in this restructuring has reached for the same vocabulary: reskill, upskill, future-ready. TCS introduced a bench policy under which an engineer unassigned to a project for 35 days must be retrained or exited. The company claims over 65% of its bench force was realigned to new digital projects. Wipro trained 2.35 lakh employees in AI basics, with 50,000 completing advanced modules. NASSCOM’s FutureSkills programme aims to train four lakh workers. TCS is training 25,000 engineers on Microsoft Azure OpenAI tools.
These numbers are not false. They are, however, incomplete. Forrester Research’s Predictions 2026 report found that only 23% of AI decision makers surveyed said their organisations had offered prompt engineering training in 2025. Employees are mostly teaching themselves. More telling is Forrester’s finding that 55% of employers already regret AI-driven layoffs. The report predicts that half of AI-attributed job cuts will be quietly rehired, but offshore or at significantly lower salaries. The Swedish fintech firm Klarna replaced 700 customer service employees with AI, saw quality decline, faced a customer revolt, and was subsequently compelled to rehire humans. That case is not an outlier; it is an early warning about the gap between AI promise and AI performance in complex, judgment intensive work.
The deeper problem is structural. Geoffrey Hinton, the Nobel Prize winning computer scientist widely described as the ‘godfather of AI,’ has warned that AI will increase unemployment while driving higher profits, a consequence he attributes to capitalism’s incentive structure rather than to the technology itself. Hinton noted that mass layoffs have not yet fully materialised, but that AI is already reducing entry level opportunity – the rung where careers are built and where character is formed on the job. You cannot retrain someone for a role they never had the chance to hold in the first place.
India’s IT sector built the country’s urban middle class not through charity or policy, but through the ruthless scalability of a labour model that happened to be aligned with global demand. That model is now structurally misaligned. The firms that will survive this transition are those reckoning honestly with what AI can actually do today, not what it might do in five years, and investing in the human capabilities that no algorithm has yet been trained to replace: complex reasoning, accountability, cultural judgement, and the capacity to say, “this is wrong” when a client insists otherwise. The freshers sitting in Bengaluru’s engineering colleges right now will build their careers in an industry remade by the very technology they are being asked to learn. They deserve an honest account of the terrain ahead, not a jugaad reassurance that everything will be fine. History suggests it will be different. Whether it will be better depends on choices being made right now, mostly in boardrooms, and mostly without them in the room.
Summary
- TCS cut 12,261 employees in July 2025, representing 2% of its global workforce, its largest layoff since FY2015, driven primarily by skill mismatches as AI restructures what clients demand.
- India’s IT industry, which employs 5.67 million people and contributes over 7% to GDP, built the country’s urban middle class on a labour-arbitrage model that AI is now systematically eroding.
- Globally, nearly 55,000 US job cuts were attributed directly to AI in 2025 (Challenger, Gray & Christmas), while Amazon, Microsoft, and Salesforce each cited AI explicitly when announcing major workforce reductions.
- The entry-level hiring collapse is a separate, compounding crisis: the Burning Glass Institute found the share of entry-level software jobs dropped from 43% to 28% between 2018 and 2024, cutting off the apprenticeship pipeline that built careers for a generation.
- Forrester Research found that 55% of employers regret AI-driven layoffs, and predict half will rehire quietly at lower salaries or offshore, suggesting many firms are acting on AI’s promise rather than its proven performance.
- Infosys has deployed over 300 enterprise AI agents delivering 5–15% productivity gains; Wipro has rolled out 200+ AI agents in internal functions; both companies are growing revenue without proportionate headcount growth.
- The reskilling response, from TCS’s 35-day bench policy to NASSCOM’s FutureSkills programme targeting four lakh workers, is real but insufficient: only 23% of AI decision-makers (Forrester) offered prompt engineering training in 2025.
- India produces 1.5 million engineering graduates annually; without meaningful entry-level absorption, the social dividend of the IT boom will not pass to the next generation.
- The WEF’s Future of Jobs Report 2025 projects a net global job loss of 14 million by 2027, with 83 million eliminated and 69 million created, a chasm that reskilling rhetoric alone cannot bridge.
Subscribe Deshwale on YouTube
