AI could erode India's outsourcing advantage, warns WB report - Rediff.com

TEXT START: Generative artificial intelligence is already reshaping South Asia's job market, with multinational companies and firms integrated into global value chains reducing recruitment more sharply than domestic firms, according to the World Bank's World Development Report 2026.

The Dissection

The article is a controlled warning. It admits that AI is already reducing recruitment in precisely the firms most connected to India's export economy, reports a 39% decline in developing-country freelance outsourcing in 2025, and identifies call centres, customer support, data processing, and entry-level knowledge work as exposed. Then it dilutes the diagnosis with reassuring percentages, productivity anecdotes, and advice to adopt foreign-built tools locally.

The text is not describing a temporary hiring cycle. It is documenting the first visible fracture in India's labour-arbitrage model: firms with the greatest access to AI are cutting the human layer that made outsourcing profitable.

The Core Fallacy

The central error is confusing low current automation exposure with durable economic safety.

“Fewer than one in 10 jobs” being susceptible to automation does not protect the workers concentrated in the exposed sectors. AI does not need to automate most jobs to destroy the wage ladder. It only needs to eliminate enough routine, entry-level, and internationally tradable work to collapse hiring, depress wages, and block advancement into the middle class.

The article also treats AI enhancement as if productivity gains automatically create employment. Under Discontinuity Thesis mechanics, higher productivity can mean fewer workers producing the same output. The gains accrue to the owners and controllers of AI capital; the displaced worker receives neither productive participation nor an automatic claim on the surplus.

The deeper omission is ownership. India may adapt AI to local languages and sectors, but adaptation does not make India sovereign. If the models, chips, cloud infrastructure, capital, and platforms remain controlled abroad, India is improving its productivity while renting the machinery of its own labour displacement.

Hidden Assumptions

• Export demand will continue expanding fast enough to absorb workers made redundant by AI.
• Productivity gains will translate into wages rather than higher margins and reduced headcount.
• New AI-related jobs will appear at the scale and skill distribution needed to replace routine service work.
• Domestic firms’ slower adoption is a protective moat rather than delayed exposure.
• Outsourcing can remain competitive when the client can automate or relocate the task entirely.
• Local adaptation of imported systems creates development rather than deeper technological dependence.
• The agricultural example can be generalized to the millions dependent on digitally delivered services.
• Governments can redistribute AI-generated surplus without confronting the concentrated ownership of the infrastructure.
• “Enhanced” workers retain bargaining power after AI makes their output easier to reproduce and compare.

None of these assumptions is established by the evidence presented. Several run directly against it.

Social Function

Primary classification: transition management. Secondary classifications: partial truth and ideological anesthetic.

The article prepares the public for declining employment while preserving faith in the existing development script. Its message is: the outsourcing engine is being cannibalized, but adoption, localization, and better policy can keep the machine running. That is useful governance language, not a solution to the ownership problem.

The warning is real. The anesthesia is the claim that aggregate exposure rates and scattered productivity gains meaningfully answer the employment question. They do not. A farmer saving money through better forecasts does not restore the mass wage-and-consumption circuit being severed in software and business services.

The Verdict

The report captures the opening movement of the discontinuity but refuses to follow it to its structural conclusion. India’s outsourcing advantage was built on the price and availability of human cognitive labour; AI attacks both foundations simultaneously. The immediate casualty is entry-level work, but the strategic casualty is the ladder that converted low-cost service labour into middle-class employment.

Domestic firms’ delay is hospice care, not protection. Local AI adaptation may raise output, but without sovereign control of AI capital it can accelerate dependence while shrinking the labour share of income. Under P1, P2, and P3, India is not facing a minor erosion of comparative advantage. It is facing the collapse of the comparative advantage itself—and the loss of a promised route to mass productive participation.