Uneven AI adoption risks widening capability inequality: low-income groups and segmented labour markets—notably in India’s IT services—face accelerated skill-change and transition risks, calling for coordinated policies on learning equity, transition protections and accountable algorithmic management.

Main Finding

AI adoption is accelerating in advanced economies but access to complementary capabilities (knowledge, digital infrastructure, compute, organizational adoption) is uneven. This widening “capability inequality” risks producing a new, distributionally biased phase of growth—what international bodies call a possible “Next Great Divergence”—where gains from AI concentrate unevenly across countries, firms, and worker groups. In labor markets the dominant pattern is task transformation and accelerated skill change, not uniform job destruction, creating acute transition risks for vulnerable workers and for sectors (notably Indian IT services) that have been key mobility channels.

Key Points

• Uneven adoption and capability gaps:
  • Firms in advanced economies show rapid AI uptake while many countries, regions, and firms lag (OECD 2026; UNDP 2025; WTO 2025; WEF).
  • “Capability inequality” captures unequal access to knowledge, digital infrastructure, compute resources, and organizational readiness—these shape who can realize AI productivity gains.
• Labor-market effects:
  • Refined exposure measures point to widespread task transformation and accelerated skill change rather than uniform mass job loss (ILO 2025).
  • Vulnerable workers face higher reskilling burdens and socio-economic insecurity due to limited access to learning and digital resources.
• India-specific dynamics:
  • Macro growth remains robust, but labor-market segmentation, informal work, and digital capability gaps produce distributional vulnerabilities (MoSPI–NSO 2025; PLFS; HCES).
  • AI-driven efficiency pressures in IT services could compress billable work, change hiring and wage structures, and raise transition risks even for technical workers (Reuters 2026a; Nasscom).
• Policy response proposed:
  • A “shared gains” architecture emphasizing learning equity (universal access to reskilling and lifelong learning), transition protections (income support, retraining, active labor programs), accountable algorithmic management, and distribution-sensitive metrics beyond GDP.

Data & Methods

• Conceptual framework:
  • Integrative socio-technical political economy approach that links technological capability, institutional context, and distributional outcomes across past, present, and forecasted future phases.
• Sources and validation:
  • International reports: OECD (2026), ILO (2025), UNDP (2025), WTO (2025), WEF.
  • Official Indian statistics: MoSPI/NSO GDP estimates, Periodic Labour Force Survey (PLFS), Household Consumption & Expenditure Survey (HCES).
  • High-reliability sector evidence: Reuters reporting, Nasscom analyses.
• Analytical structure:
  • Synthesis of secondary validated sources combined with official statistics.
  • Temporal framing across historical adoption, current diffusion patterns, and projected future impacts.
  • Use of refined exposure measures (ILO) to assess task-level transformation rather than coarse occupation-level displacement.

Implications for AI Economics

• Research priorities:
  • Move beyond GDP-centric analysis to measure capability inequality (access to compute, data, skills, institutional adoption) and its distributional consequences.
  • Develop micro–macro models that incorporate firm heterogeneity in AI adoption, complementarities between AI and skills, and task-level labor reallocation.
  • Improve measurement: firm‑level adoption surveys, matched employer–employee panels, administrative data on transitions, and task-level exposure indices.
• Policy evaluation and design:
  • Evaluate policies that expand learning equity (public reskilling, access to computing and digital infrastructure) and test targeted transition protections (wage insurance, retraining vouchers).
  • Incorporate algorithmic accountability and workplace governance into labor-market regulations and collective bargaining frameworks.
  • Design cross-country and within-country measures to monitor a “Next Great Divergence” and trigger international cooperation where capability gaps threaten global inequality.
• Modeling and empirical strategies:
  • Use natural experiments and RCTs to test upskilling programs and transition supports; employ structural/CGEmodeling to assess long-run distributional effects under varying adoption scenarios.
  • Analyze sectoral channels (e.g., IT services in India) as case studies for mobility impacts and wage/billable-hour dynamics.
• Policy relevance for economists:
  • Recognize institutional context and access to complementary resources as first-order determinants of AI’s distributional outcomes.
  • Prioritize metrics that capture socio-economic security (income volatility, transition duration, access to reskilling) alongside productivity gains.