AI-driven change in India is widening labour-market divides: high-skill, high-wage roles are growing while routine middle-skill jobs shrink, and weak reskilling efforts risk leaving many workers stuck in low-skill employment.

Main Finding

AI adoption in India is producing pronounced labour-market polarization: high-skilled, high-wage occupations expand and capture most gains, routine middle-skill jobs shrink, and a large low-skill segment persists. Limited reskilling coverage and institutional capacity risk trapping many workers in low-skill employment and widening wage inequality. Comparative evidence from the UK shows that systematic reskilling and coordinated adoption can reduce displacement and ease transitions, offering policy lessons for India.

Key Points

• Theoretical framing: interprets observed changes through Skill-Biased Technological Change (SBTC) and Human Capital Theory — AI complements high-skilled work and substitutes routine middle-skill tasks, while returns to skills rise.
• Polarisation pattern in India: simultaneous growth at the top, contraction of the middle, and persistence of a sizable low-skill workforce — distinct from some advanced-economy patterns where middle-skill erosion can be partially offset by upskilling.
• Reskilling gap: existing training programs and coverage are insufficient in scale, accessibility and alignment with employer needs, limiting labour mobility into growing high-skill roles.
• Comparative insight: the UK’s more systematic AI adoption and structured training ecosystem reduces displacement risk and provides transferable lessons (coordination, targeted programs, quality assurance).
• Risk of a “low-skill trap”: without scaling effective reskilling and inclusive policies, AI could entrench inequality and underemployment in emerging economies.

Data & Methods

• Data sources: analysis uses secondary data and official reports from 2020–2024 (labour and employment statistics, government and industry reports, training program evaluations and related secondary sources).
• Methods: theory-driven empirical synthesis combining SBTC and human-capital frameworks with descriptive trend analysis and cross-country (India–UK) comparison to draw policy-relevant inferences.
• Evidence strategy: documents sectoral and occupational shifts, wage and employment trends across skill groups, and evaluates institutional capacity for reskilling using available official and secondary assessments.
• Limitations: reliance on secondary sources constrains causal identification; heterogeneity across regions, sectors and informal employment requires careful interpretation.

Implications for AI Economics

• Distributional modeling: AI economics must incorporate asymmetric complementarities (AI×skill) and frictions to upskilling when estimating welfare and labour-market impacts.
• Policy design priorities:
  • Scale targeted reskilling and upskilling programs aligned to employer demand and regional labor-market needs.
  • Strengthen public–private partnerships to co-design curricula, apprenticeship pathways and certification systems.
  • Invest in early-stage education and lifelong learning infrastructure to raise baseline human capital and adaptability.
  • Implement active labour-market policies (job-search support, portable credentials, transition subsidies) to lower frictions.
  • Ensure social protection for displaced workers while facilitating retraining and job transitions.
• Research agenda: quantify the costs and returns of alternative reskilling models, study heterogeneity across informal vs formal sectors, and evaluate policies that most effectively prevent a low-skill trap.
• International lesson transfer: emerging economies can adapt institutional elements from advanced-economy models (e.g., coordinated training pipelines, quality assurance, employer co-financing) but must tailor programs to local labour-market structures and resource constraints.