A landmark NASSCOM-Deloitte study on artificial intelligence's impact on India's IT sector workforce — released in May 2026 after 18 months of primary research across 150 companies — found that AI automation had eliminated approximately 52,000 IT jobs in India over the prior two years, primarily in categories including basic software testing, data entry and validation, routine code review, tier-1 customer support and standardised report generation, while simultaneously creating approximately 78,000 new technology jobs in categories including AI/ML engineering, prompt engineering, AI governance and ethics, large language model fine-tuning, AI product management and data science for AI training. The net positive job creation figure of 26,000 roles was characterised by the authors as an underestimate of total displacement, since many workers displaced from eliminated roles have not yet found comparable new employment, creating a transition period of structural unemployment in specific IT skill categories.
The displacement is concentrated in specific workforce segments rather than being broadly distributed across the IT sector. Entry-level engineers and freshers with less than two years of experience doing routine code implementation, testing and debugging have been the most affected group, as AI coding assistants like GitHub Copilot, Amazon CodeWhisperer and Google's Gemini for Code can now perform many tasks that were previously the domain of junior developers — reducing the size of fresher batch requirements at major IT services companies. TCS, Infosys and Wipro collectively hired 82,000 freshers in FY26 versus 220,000 in FY22 — a 63% reduction that reflects both cyclical IT services demand moderation and the structural productivity improvement from AI tools that allows existing engineers to deliver significantly more output per person. The career entry ramp into IT services, which has been the primary vehicle for upward mobility from engineering colleges to the middle class for millions of Indians, has consequently become significantly more competitive and skill-intensive.
The new roles being created by AI adoption are substantially different from the eliminated roles in terms of required skills, compensation levels and career trajectory. An AI product manager at a leading IT company earns Rs 25-45 lakh annually — 2-3x the salary of the junior developer roles being displaced — reflecting the significantly higher skill requirements and the genuine value created by effective AI implementation. However, the skilling gap between displaced workers and the requirements of new roles is enormous: basic software testing requires 6-month reskilling for most workers, while effective AI/ML engineering requires 18-36 months of deep technical retraining that is beyond the reach of many workers with families, financial obligations and limited formal education in the underlying mathematics of machine learning.
India's IT sector response to the AI-driven workforce transition has been both self-interested and genuinely constructive. The major IT companies have all launched large-scale AI retraining programmes — Infosys's Lex platform has trained 280,000 employees in AI tools, TCS's AI-powered learning platform has completed 4.2 crore course completions across its global workforce, and Wipro has partnered with Coursera for 100,000 AI certification completions annually. However, critics note that these initiatives primarily serve the companies' existing workforce of experienced engineers who are being upskilled to use AI tools in their current roles — rather than addressing the more difficult challenge of retraining displaced entry-level workers or the 8 lakh+ engineering graduates who emerge from India's colleges annually and face a fundamentally changed job market compared to what their education prepared them for.
The policy response to AI-driven displacement has been notable for its relative under-action compared to the scale of the challenge. While NASSCOM and the government have established various skilling programmes, the speed and scale of AI-driven workforce transformation is outpacing the capacity of training institutions to respond. The most constructive emerging model is industry-academia partnership where companies like TCS, Infosys and Google are co-designing curricula with engineering colleges to ensure that graduates enter the job market with genuine AI proficiency rather than outdated programming skills. The government's AI Mission, with its dedicated compute infrastructure and training programmes, is also expanding rapidly. But the deepest challenge — ensuring that India's 80 lakh + engineering graduates and the much larger cohort of IT sector workers facing displacement can navigate the AI transition to productive, dignified employment — will require sustained, coordinated action from government, industry, educational institutions and civil society over the next decade that has not yet been fully mobilised at the required scale and urgency.
