India’s AI adoption trajectory is accelerating across services, manufacturing, and platform-based work. Firms are deploying automation to improve productivity, reduce costs, and remain globally competitive. However, India’s labour market, dominated by informality, low skill depth, and uneven sectoral transformation, raises a critical concern: will AI-led productivity gains translate into commensurate employment absorption?

This article examines whether India risks a phase of automation without absorption, where output and efficiency rise but labour demand fails to adjust, threatening employment outcomes and India’s demographic dividend.

The analysis focuses on four interlinked issues: task displacement versus job creation, sectoral mismatch, skill constraints, and long-term demographic risk.

How the Risk Materialises: Automation Without Absorption

1. Task Displacement Concentrated in Employment-Generating Segments

Quantitative evidence suggests that AI exposure in India is uneven but increasingly consequential. The IMF estimates that 26% of Indian jobs are exposed to AI, with roughly 12% facing high substitution risk and the remaining exposed jobs potentially experiencing task augmentation rather than full displacement . While this exposure is lower than in advanced economies due to India’s large agricultural and informal workforce, it is concentrated in sectors that drive formal employment and wage growth.

White-collar labour market data show early displacement effects. A World Bank analysis of online job postings across South Asia found that since the widespread adoption of generative AI (2022 onward), job listings for highly AI-exposed occupations declined by approximately 20%, particularly in business services, IT support, and clerical roles . Over the same period, postings for AI-specialist roles increased sharply but remained a small base, around 6-7% of white-collar job listings, which are insufficient to offset absolute losses in routine professional roles.

India’s IT-BPM sector illustrates this imbalance clearly. The sector employs ~5.4 million workers and contributes 7-8% of GDP, but automation is directly reducing demand for mid-skill roles such as testing, maintenance, and back-office processing. Reuters reports that large firms have already begun workforce rationalisation, with industry estimates suggesting 300,000-500,000 outsourcing jobs at medium risk over the medium term due to AI deployment. Job creation in AI engineering and data science remains skill-intensive and limited in scale.

This asymmetry highlights a core issue: AI creates jobs at the top of the skill distribution while displacing tasks in the middle, leading to net productivity gains without proportional employment expansion.

2. Sectoral Elasticity Determines Whether Productivity Converts into Jobs

India’s current economic trajectory is defined by a significant structural imbalance. While the services sector drives over 55% of GDP, it absorbs only 30% of the workforce. Conversely, agriculture remains a low-productivity “labour sink,” employing nearly 46% of workers but contributing a mere 20% to GDP. Despite long-standing industrial initiatives, manufacturing and construction have failed to significantly bridge this gap, stagnating at 11.4% and 12% of employment respectively as of 2024.

Analysis from Business Standard reveals a “dual divergence” within the service economy. While the aggregate employment elasticity in services rose to 0.63 post-COVID, this headline figure masks deeper structural fragilities:

• The High-Value Peak: Modern segments like IT, Finance, and Healthcare show high elasticities (0.88–0.95), meaning they hire rapidly when they grow. However, these sectors employ a very small fraction of the total workforce and require specialized skills that fewer than 10% of engineering graduates currently possess.
• The Labour-Intensive Base: Traditional sub-sectors such as trade and personal services, which employ the vast majority, show a declining responsiveness to growth. In capital-heavy pockets like telecommunications and insurance, elasticities have actually turned negative, suggesting that “capital deepening” and automation are displacing workers rather than creating new roles.

RBI national accounts data further underscores this disconnect. Between FY15 and FY23, services output grew at a 6% CAGR, yet employment growth lagged at just ~2%.

AI adoption reinforces this pattern. Productivity gains accrue to sectors already weak in employment generation, such as IT services, finance, logistics platforms, while labour-intensive sectors such as textiles, food processing, and MSME manufacturing face slower AI diffusion and tighter margins. Only 3% of Indian enterprises possess sufficient in-house talent and resources to make the most of what AI has to offer, while the remaining 97% of executives cite lack of talent as a primary hurdle, with 23% reporting that they’re in no state of data readiness to take up AI deployments.

As a result, India risks a dual economy: a high-productivity, AI-enabled formal sector with limited hiring, and a large informal sector where employment persists but productivity stagnates.

3. Firm-Level Constraints Prevent Broad-Based AI Complementarity

Enterprise-level data further explains why AI adoption does not translate into widespread employment effects. Only 3 % of Indian enterprises possess sufficient in-house capability to deploy AI effectively. Among the remaining 97 %, lack of skilled talent is the primary constraint, and 23 % report insufficient data readiness.

This creates a bifurcation:

• Large, capital-intensive firms deploy AI to raise efficiency but do not scale hiring.
• MSMEs and labour-intensive firms lack the capacity to adopt AI in ways that would raise productivity and wages.

The result is not technological unemployment in aggregate, but employment stagnation in productivity-enhancing segments, reinforcing dualism in the labour market.

Why Skill Gaps Slow Labour Reallocation

The reallocation of labour from declining tasks to emerging roles is constrained primarily by skills. PLFS data 2023-24 consistently show that 4.1% of people aged 15-59 years have received formal vocational/technical training, while another 30.6 per cent received training through informal sources.

The World Economic Forum estimates that 38% of skills used in India’s workforce will change by 2030, one of the highest transition rates globally. Yet skilling capacity remains inadequate. Ministry of Skill Development data indicate annual training capacity of ~15-20 million candidates, while the labour force adds 8-10 million new entrants annually, not accounting for reskilling needs of existing workers.

According to the World Economic Forum’s Future of Jobs 2025 report, about 40% of Indian firms expect an inability to attract the right talent for their operations, and 65% identify skill gaps as a major constraint on future growth; a finding that highlights both employer demand pressures and worker skill deficits. This aligns with broader market signals: India Skills Report 2026 shows that despite employability rising to about 56.35% overall, deeper technical readiness, particularly in AI, data analytics, and advanced computing, remains concentrated in a relatively small cohort of workers, leaving the bulk of graduates and mid-career professionals inadequate for the skills required by emerging AI-enabled industries.

This solidifies that skill shortages are systemic and tied to both education-to-employment barriers and employer hiring constraints, further creating frictions that slow labour reallocation. Workers displaced from routine service roles cannot easily transition into AI-complementary jobs, leading to underemployment rather than smooth occupational mobility.

When the Risk Becomes a Demographic Problem

India’s demographic dividend depends not merely on workforce size but on productive employment creation. With over 65% of the working-age population, India requires sustained job creation at scale. However, PLFS data show that while labour force participation has improved, much of the employment growth since 2020 has occurred in self-employment and informal work, not in high-productivity formal jobs.

Economists have flagged that official unemployment rates mask underemployment and low-quality job creation. World Bank estimates suggest that nearly 70% of jobs in India are vulnerable to automation over the long term, especially informal and routine roles. If AI adoption accelerates without matching absorption channels, the demographic dividend risks erosion through stagnant wages, rising informalisation, and youth underemployment.

The Economic Survey cautions that technological transitions without parallel human capital investment can widen inequality and slow consumption growth, undermining the very demand conditions needed for job creation.

Conditions Under Which the Risk Can Be Mitigated (Derived from the Evidence)

Based strictly on the data presented, AI-led employment outcomes improve only when three conditions are simultaneously met:

1. High Employment Elasticity in AI-Adopting Sectors: Productivity gains must occur in sectors capable of absorbing labour. Current data show this condition is weak in India’s AI-leading sectors.
2. Scalable Skill Transition Capacity: With only 4.1 % formally trained workers and limited reskilling throughput, labour mobility remains constrained. Without expanding skill depth, AI complements capital, not labour.
3. Broad-Based Firm Readiness: With only 3 % of firms AI-ready, productivity gains remain concentrated. Employment effects require diffusion beyond large firms into MSMEs and labour-intensive segments.

Absent these conditions, the empirical pattern observed—automation without absorption—is likely to persist.

Aligning AI Adoption with Labour Absorption is a Structural Challenge

India’s AI transition is not constrained by technology but by institutional and structural factors. The data show that productivity gains are materialising faster than employment absorption due to sectoral rigidity, skill constraints, and firm-level readiness gaps.

AI-led growth can coexist with employment expansion only if productivity gains shift toward high-elasticity sectors, skill transitions scale beyond current capacity, and AI adoption diffuses across firm sizes. Without these conditions, AI will deepen structural dualism rather than resolve it.

Protecting India’s employment outcomes and its demographic dividend requires aligning technological adoption with labour-absorbing growth mechanisms, not assuming that productivity gains will translate automatically into jobs.

At Tatvita Analysts, we focus on how risks materialise within real economic systems and the conditions under which they can be governed. India’s AI transition demonstrates that automation outcomes are shaped less by technology itself and more by sectoral structure, skill depth, and institutional readiness.