Artificial intelligence is no longer merely a technological story. It is becoming a labour-market story, a social-stability story and, increasingly, a middle-class survival story. For nearly three decades, countries such as India benefited from a historic global shift in white-collar work. Multinational firms outsourced software development, back-office processing, accounting support, customer service and research functions to a vast English-speaking workforce.
Entire cities were transformed by this model. Bengaluru, Hyderabad, Pune and Gurgaon emerged as symbols of aspirational India precisely because knowledge work appeared immune to the disruptions that had earlier devastated factory labour. That assumption is now under strain. The first wave of artificial intelligence largely functioned as an assistant. It helped workers write code faster, summarise documents, analyse data and automate repetitive tasks. Companies presented these tools as productivity enhancers rather than job destroyers. But the next stage of AI development is qualitatively different. “Agentic AI” ~ systems capable of independently executing complex tasks with minimal human supervision ~ threatens to alter the economics of white-collar employment itself.
The implications extend far beyond the technology sector. Across advanced economies, firms are already slowing recruitment for junior roles. Entry-level programming, technical writing, basic legal drafting and standardised analytical work are increasingly being handled by AI-assisted systems. Corporations may still require experienced professionals to supervise outcomes, but the pyramid beneath them is narrowing. That matters because the middle class depends not merely on elite jobs, but on a steady pipeline of entry-level employment that allows younger workers to enter the system. India is especially exposed because its service economy was built around the kind of structured, rules-based and English-language tasks that AI performs well.
The country’s IT-services industry succeeded because it industrialised knowledge work. Artificial intelligence may now industrialise it further ~ with fewer humans required in the chain. The danger is not immediate mass unemployment. AI systems still suffer from reliability problems. Autonomous decision-making remains risky in fields where errors carry financial, legal or reputational costs. High-end computing infrastructure is also expensive, limiting large-scale deployment. But technological history suggests these constraints rarely remain permanent.
Costs fall, systems improve and businesses adapt rapidly once efficiency gains become undeniable. The deeper risk lies in the social consequences of a prolonged squeeze on educated salaried workers. India’s urban middle class is already burdened by rising housing costs, stagnant wages, consumer debt and shrinking job security. If white-collar employment loses the stability it once promised, the economic anxieties currently visible in parts of the West could become sharper in India’s densely competitive urban centres. Governments have not yet grasped the scale of the transition. Public debate still treats AI largely as an innovation story rather than a structural labour disruption. Reskilling alone may not be enough if technology begins reducing the overall demand for large sections of routine knowledge work. The AI revolution will undoubtedly create wealth. The unanswered question is whether it will also create enough human work.