AI Is Taking Jobs — But India Could Be the Biggest Winner of the Artificial Intelligence Boom
Newgen Software shares rise 12% on AI push! (Image X.com)
By SAHASRANSHU DASH
India’s experience suggests artificial intelligence is reshaping employment rather than simply eliminating jobs, while AI-enabled services could open a new export pathway for the world’s fifth-largest economy.
Sheffield (UK), September 7, 2026 — The widespread doom-mongering and dire predictions that artificial intelligence will trigger mass unemployment sit uneasily with the evidence emerging from India. A report by Nomura in August 2026 suggests that, rather than destroying jobs outright, AI is reshaping the composition of employment.
Between 2022 and August 2026, India recorded approximately 83,100 AI-related hires, compared with around 31,900 layoffs or role changes directly attributable to AI adoption. In other words, roughly 2.6 AI-enabled jobs have been created for every worker displaced. (Nomura argues that India is the world’s most important test case because its economy is unusually dependent on information technology services and business-process outsourcing.) So far, the data point towards labour market transformation rather than technological unemployment.
This should not be entirely surprising. Economists have long rejected the ‘lump of labour’ fallacy, i.e. the belief that there is a fixed amount of work in an economy, so every machine must permanently replace a worker. History suggests otherwise. Mechanisation reduced agricultural employment but created manufacturing industries. Computers eliminated typists while generating software engineers, cybersecurity specialists and digital marketers. Technology rarely leaves labour markets untouched, but it generally changes what people do (in other words, the nature of work) rather than eliminating the need for human labour altogether.
This insight underpins the task-based approach to technological change developed by Daron Acemoglu, David Autor and Pascual Restrepo. Their research argues that technology automates tasks, not occupations. Whether employment rises or falls depends on whether automation simply substitutes for existing labour or simultaneously creates new tasks in which humans retain a comparative advantage. Historically, it has been this second mechanism that has prevented long-run technological unemployment.
Acemoglu’s more recent work on generative AI reinforces this point. In The Simple Macroeconomics of AI, published in May 2024, some months before he was awarded the Nobel Prize in Economic Sciences, he developed a task-based model in which AI affects the economy through both automation and human-task complementarities. Using existing estimates of AI exposure and task-level productivity improvements, he concluded that automation alone would raise total factor productivity (TFP) by no more than 0.66% over a decade, and probably less than 0.53%, because much of the early evidence comes from relatively simple tasks that are easier for AI to master than the more complex, context-dependent work that dominates many professions.
Aggregate productivity, therefore, will increase only modestly unless firms redesign production around human-AI complementarities and create entirely new labour-intensive activities. Automation alone substitutes merely capital for labour, but genuine productivity revolutions occur when technology expands the range of economically valuable work.
At what stage will robots become conscious, asks Karan Singh
From Routine Coding to Knowledge Services
India’s experience so far appears remarkably consistent with this interpretation. Now, there are certainly adjustment costs. The traditional labour-arbitrage model that underpinned the growth of firms such as TCS, Infosys and Wipro, with large numbers of junior programmers performing routine coding, testing and maintenance, is under increasing pressure. Large language models now generate boilerplate code, documentation and test cases efficiently, and in seconds. Entry-level recruitment has slowed as routine programming becomes increasingly automated.
Yet focusing solely on declining graduate recruitment overlooks the much larger structural transformation under way. AI is simultaneously creating demand for a new generation of higher-value knowledge-intensive services. Rather than exporting inexpensive coding capacity alone, Indian professionals increasingly deliver end-to-end AI-enabled business solutions: enterprise workflow automation, AI agents, cybersecurity, financial analytics, tax technology, healthcare diagnostics, legal technology, supply-chain optimisation and AI governance.
‘Vibe coding’ enables small teams to build products that previously required dozens of software developers. Competitive advantage therefore shifts from writing every line of code manually towards understanding business processes, exercising judgement and integrating technical expertise with commercial knowledge.
This pattern is increasingly supported by empirical research. A landmark field experiment by Erik Brynjolfsson, Danielle Li and Lindsey Raymond found that customer-service agents using generative AI became approximately 15 per cent more productive, with the largest gains accruing to less experienced workers. AI therefore acted less as a substitute for labour than as a mechanism for diffusing expertise across the workforce. Similarly, research by Acemoglu, Autor et al finds that firms adopting AI increasingly alter skill requirements and recruit workers capable of complementing AI systems, while economy-wide evidence of large-scale job destruction remains remarkably limited.
AI and India’s New Export Model
Indeed, AI may offer India something even more significant than higher productivity. It may provide an alternative path to export-led development at a time when the traditional manufacturing model faces unprecedented constraints.
China’s industrial rise occurred during the era of hyperglobalisation, when tariffs were falling, multinational firms were relocating production and global supply chains were expanding rapidly. India enters the global economy under almost the opposite conditions: slower world trade growth, protectionism, industrial policy in advanced economies, friend-shoring and, perhaps most importantly, persistent Chinese industrial overcapacity. Chinese producers now dominate many manufacturing sectors through enormous economies of scale, integrated supply chains and accumulated learning, making it far more difficult for late industrialisers to replicate China’s export miracle.
Knowledge-intensive services are fundamentally different. Unlike manufacturing, exporting AI-enabled services requires relatively little physical capital. The principal inputs are human capital, digital infrastructure and computing capacity. AI further reduces the marginal cost of exporting expertise across borders. A software engineer in Bengaluru can build an AI-enabled CRM platform for a Swedish manufacturer, a transfer pricing specialist in Bhubaneswar can advise multinational clients worldwide, a digital marketer in Kochi can manage campaigns for a Filipino retailer, and a doctor in Chennai can combine AI diagnostics with international telemedicine. Rather than replacing skilled professionals, AI enables them to serve far more clients simultaneously.
The macroeconomic implications could be substantial. Goldman Sachs projects India’s services exports reaching around US$800 billion by 2030, while the Economic Survey 2025–26 finds that AI-intensive service sectors have experienced approximately 39.5 per cent faster export growth than less AI-intensive services since the diffusion of generative AI. The WTO similarly estimates that AI could increase global trade by more than one-third by 2040, with digitally deliverable services expanding even faster as information and transaction costs decline.
This is precisely the kind of development Acemoglu argues should generate sustained productivity growth. AI is not merely replacing existing tasks; it is creating entirely new exportable services built around human expertise enhanced by intelligent machines. For a country possessing one of the world’s largest pools of English-speaking engineers, accountants, lawyers, doctors and management professionals, this represents a significant comparative advantage.
None of this implies that manufacturing has become unimportant. India still requires a strong industrial base to generate employment for workers without advanced qualifications, support technological learning and strengthen strategic resilience. AI cannot absorb the millions of workers who might otherwise have entered labour-intensive manufacturing. Nevertheless, it may allow India to compensate for some of the limitations imposed by a more fragmented global trading system.
Rather than attempting to replicate China’s manufacturing-led development model under far less favourable international conditions, India may become the first major economy to achieve sustained convergence through a combination of advanced manufacturing and globally scalable AI-enabled services. Once that transition ripens, India’s comparative advantage will increasingly lie not in exporting inexpensive labour embodied in manufactured goods, but in exporting highly productive human capital embodied in digitally deliverable professional services. Far from heralding the end of India’s growth story, artificial intelligence will become one of its most important engines.
The Code, Calm and Caution: India Inc. Sizes Up the AI Moment
Follow The Raisina Hills on WhatsApp, Instagram, YouTube, Facebook, and LinkedIn
About The Author
