The Ground Is Already Shifting
Walk into any training center in Hyderabad's Ameerpet neighborhood today. The courses on the board used to say Java and web development. Now they say AI data science and prompt engineering. The nine-month course costs over Rs 1.1 lakh - more than double what a traditional coding course used to cost. Recruiters stopped asking for the old skills. The shift is already here, visible in every hiring dashboard and every emptied training seat.
India built its economic identity on a simple idea: millions of smart, English-speaking graduates who would do the work that Western companies needed done, at a fraction of Western salaries. That idea made India the back office of the world. It created a middle class in Bengaluru, Hyderabad, Pune, and Gurugram. It funded apartments, schools, and restaurants. It gave millions of families a first step into financial security.
AI is now attacking that idea at its foundation.
The Scale of the Problem
India's IT and tech services sector employs about 7.5 to 8 million people. According to NITI Aayog's October report - developed with NASSCOM and BCG - headcount in that sector could fall to 6 million by the end of this decade if nothing changes. That is roughly 2 million jobs gone.
The customer service and BPO sector employs another 1.65 million people. NITI Aayog projects that number could fall from 2 to 2.5 million to 1.8 million. Every number in that range represents a real family.
NITI Aayog CEO BVR Subrahmanyam put it directly: "Don't look at it as just 2 million jobs - those 2 million probably support an ecosystem of 20 to 30 million others."
TCS, India's largest IT company, cut its headcount to around 580,000 - down more than 20,000 from its peak. The company had hired 100,000 people in a single year just before that. Hiring at Infosys has pulled back sharply. Indian IT stocks fell roughly 20 percent as investors reacted to what they saw coming.
The threat is not theoretical. A Bengaluru startup called LimeChat built AI agents that handle up to 95 percent of customer queries without any human help. Its co-founder told Reuters: "Once you hire a LimeChat agent, you never have to hire again." Over 60 percent of India's formal sector jobs are vulnerable to automation, according to the NITI Aayog report - with IT and BPO leading the exposure.
Why This Happened - and Why It Hits India Harder
India's outsourcing model had one core value: Indian developers and agents cost a fraction of what Western companies paid their own staff. Labor cost was the competitive advantage. AI has collapsed that advantage.
The marginal cost of an AI coding agent has dropped to essentially the cost of electricity. A US company that once paid for a 50-person team in Bengaluru to review legal contracts can now do it with one AI tool at almost zero cost.
The roles most at risk are entry-level and mid-level positions with low specialization. Junior developers doing repetitive code, agents answering scripted customer queries, quality assurance testers. These are exactly the roles that gave millions of graduates their first job in the formal economy.
India also has a structural gap that makes the transition harder. NITI Aayog reports that AI talent demand is growing at 25 percent per year. Supply is growing at only 15 percent. According to a Great Learning survey of engineers, 67.5 percent already feel their jobs are being negatively affected by AI - and 87.5 percent believe upskilling is now critical to survive.
India produces fewer than 500 AI-related PhDs per year. Meanwhile, 44 percent of India's top AI researchers work abroad.
What Has Already Been Tried
India has not been asleep. The government recognized the skill gap years ago and launched several programs.
FutureSkills Prime is a joint initiative between the Ministry of Electronics and Information Technology and NASSCOM. Launched in 2018, it aimed to reskill 1.4 million IT workers over five years in technologies including AI, cloud computing, and data analytics. The platform grew to over 1.85 million sign-ups, and was ranked third among 47 global digital skilling initiatives by the European Commission. India's AI skill penetration is rated at 2.8 - higher than the United States at 2.2 and Germany at 1.9, according to a BCG-NASSCOM report.
The Skill India Mission, launched in 2015, runs the Pradhan Mantri Kaushal Vikas Yojana program - a broad-based skilling initiative that covers AI and machine learning among many other sectors.
Where did these programs fall short? Scale and depth. FutureSkills Prime's 1.85 million sign-ups barely covered one-quarter of the IT workforce. The courses were often short - many just 40 to 60 hours - and lacked hands-on industry integration. NITI Aayog's own report admits the efforts remain fragmented across multiple ministries with no single coordinating body.
India built the portal. It did not build the pipeline.
How Other Countries Fixed This
Singapore - SkillsFuture
Singapore's government treats workforce retraining as national infrastructure. The SkillsFuture program, launched in 2015, gives every citizen above age 25 subsidized access to thousands of courses and job transition support. Career conversion programs, employer incentives to redesign jobs, and additional subsidies for older and disadvantaged workers are built into the structure.
Since 2016, Singapore's tech workforce agency - IMDA - has upskilled more than 340,000 individuals. In 2023 alone, about 520,000 people participated in SkillsFuture-supported programs. Singapore's tech workforce grew from 208,300 in 2023 to 214,000 even as AI disruption accelerated.
The key mechanism is mandatory employer participation. Singapore funds job redesign and pays companies to transform roles before workers are displaced - not after.
Germany - Kurzarbeit and Vocational Integration
Germany's Kurzarbeit program subsidizes wages to keep workers employed during economic disruptions and ties continued wage support to retraining requirements. Workers on reduced hours are expected to use that time for learning.
German companies are also integrating AI training directly into vocational programs before displacement occurs. The World Economic Forum highlighted Germany as a model for training workers before displacement rather than retraining them afterward. Retraining someone already unemployed is far harder than training someone who still has income and structure.
Who Is Accountable
The Ministry of Electronics and Information Technology runs FutureSkills Prime through NASSCOM. NITI Aayog has proposed a National AI Talent Mission - a unified body to coordinate across three ministries that currently run in separate silos. The India AI Mission has an outlay of Rs 10,372 crore - but most of that is focused on compute infrastructure and research, not workforce transition.
Nobody owns the answer to this question: what happens to the 2 million workers who lose IT jobs in the next five years?
What Would It Cost
NITI Aayog's roadmap projects that with the right investment, India could create up to 4 million new AI-enabled jobs in the next five years - turning a loss of 2 million into a net gain of 2 million. But that requires action now.
The Nifty IT index already shed tens of billions in investor value after AI tools began automating core outsourcing tasks. Jefferies analysts projected a worst-case scenario of 3 percent lower revenue growth for Indian IT over five years, followed by no growth beyond that.
Singapore spent over a decade and billions of dollars on SkillsFuture before it produced measurable workforce shifts. India does not have a decade.
What Needs to Happen
India has 9 million tech and customer service professionals, the world's largest pool of young digital talent, and AI skill penetration scores that already beat the US and Germany. What it lacks is execution at speed and scale.
First, India needs a single accountable body - the National AI Talent Mission proposed by NITI Aayog - with a mandate, a budget, and a deadline. One agency that owns the outcome.
Second, AI must be embedded in every engineering and computer science curriculum immediately. India has fewer than 500 AI PhDs per year. That pipeline must be rebuilt from the school level up.
Third, companies that benefit from AI automation should share the cost of transition. Companies cannot lay off 20,000 people, book the cost savings, and leave reskilling entirely to the government.
Fourth, India must stop brain drain on AI talent. Targeted return incentives - compute grid access, autonomy, research funding - would cost far less than training replacements from scratch.
The path exists. Other countries have walked it. The question is whether the government moves at the speed the problem demands - or produces another well-written report while the window closes.