Gauging Disruption: AI's Impact on Labour-Market Inequalities

Overview of LLM Usage Patterns: Recent studies by Anthropic provide valuable insights into how Large Language Models (LLMs) are being utilized globally. These studies highlight that LLM adoption is uneven across countries, with high-income nations dominating usage. Notably, the top 20 countries account for nearly 48% of per capita usage, indicating a concentration of technological access and capability.

Nature of Tasks and Adoption Trends: Initially, LLMs were primarily used for high-value specialized tasks such as coding and software development. However, as adoption widens, their application has diversified into areas like personal assistance, academic queries, and design work. This diversification has led to a marginal decline in the average economic value of tasks performed, as more routine and low-value tasks are incorporated.

Implications: The findings suggest that while LLMs are becoming more accessible, their transformative impact is still concentrated in specific sectors. For instance, developers using AI tools like Claude or GitHub Copilot significantly enhance productivity. However, broader usage patterns indicate that the technology is transitioning from a niche productivity tool to a general-purpose utility, raising questions about its long-term economic impact.

Unequal Access and Skill-Biased Technological Change: The uneven adoption of AI technologies like LLMs can exacerbate existing inequalities due to skill-biased technological change. High-income countries and skilled workers are better positioned to leverage AI, thereby increasing their productivity and income, while low-income regions and less-skilled workers lag behind.

Labour Market Polarisation: AI disproportionately benefits workers who possess complementary skills such as coding, analytical thinking, and AI integration. For example, experienced professionals who use AI tools effectively gain significantly higher productivity gains compared to entry-level workers. This creates a widening skill divide, where the benefits of AI are not evenly distributed.

Global and Domestic Implications: Globally, countries with advanced digital infrastructure gain a competitive advantage. Domestically, sectors like IT and finance may see increased productivity but also reduced demand for routine roles. For instance, customer service automation through AI chatbots can reduce entry-level hiring. Thus, without targeted policy interventions like upskilling and digital inclusion, AI adoption risks reinforcing both inter-country and intra-country inequalities.

Exposure of Occupations to AI: AI, particularly LLMs, has a significant theoretical capacity to automate tasks in occupations such as programming, financial analysis, and customer service. Studies indicate that a large proportion of tasks in these sectors can be performed by AI systems, although actual implementation remains partial.

Emerging Labour Market Trends: Contrary to fears of массов unemployment, current evidence suggests that AI has not yet led to a systematic rise in job losses. Instead, a more subtle shift is occurring: reduced hiring in entry-level roles. For example, companies are increasingly relying on AI tools to perform routine coding or data analysis tasks, thereby reducing the need for junior employees.

Case Example: In the IT sector, firms are using AI-assisted coding tools to enhance the productivity of existing employees rather than expanding workforce size. This leads to a phenomenon known as “jobless productivity growth”. Over time, this could disrupt traditional career pathways, where entry-level jobs serve as stepping stones for skill development, potentially reshaping labour market dynamics.

India’s Current Position in AI Adoption: According to the study, India ranks relatively low (98th out of 116 countries) in overall AI adoption. However, Indian users exhibit a distinct pattern of usage, focusing heavily on coding, software development, design, and academic assistance. This indicates a targeted and productivity-oriented use of AI tools.

Implications for the Workforce: This trend suggests that AI is being used as a skill and productivity multiplier in India. For instance, software engineers using AI tools can complete tasks faster and with greater accuracy. However, experienced professionals tend to benefit more due to learning-by-doing advantages, which can widen the skill gap between novice and experienced workers.

Sectoral Vulnerabilities: India’s heavy reliance on IT services, back-office operations, and routine cognitive tasks makes it particularly vulnerable to AI-driven automation. For example, AI tools can now perform tasks like code debugging or data entry, which were traditionally outsourced to Indian firms. This has already begun to reflect in pressure on IT stocks and revenue projections, highlighting the need for strategic adaptation.

Opportunities Offered by AI: AI presents significant opportunities for enhancing productivity, innovation, and global competitiveness. In India, sectors like IT and education can benefit immensely from AI integration. For example, AI-assisted learning platforms can democratize access to quality education, while AI tools can help startups scale efficiently.

Risks and Challenges: However, the risks are equally substantial. A key concern is the displacement of routine cognitive jobs, particularly in sectors where India has a comparative advantage. Additionally, the reduction in entry-level roles can disrupt the traditional skill acquisition pipeline. This could lead to a structural imbalance in the labour market, where workers lack opportunities to gain experience.

Balancing the Trade-offs: The challenge lies in managing this transition. Policymakers must focus on reskilling and upskilling initiatives, strengthening digital infrastructure, and promoting innovation ecosystems. A balanced approach that leverages AI’s benefits while mitigating its adverse effects is essential for ensuring inclusive growth. Without such measures, AI could exacerbate existing inequalities rather than alleviate them.

Scenario Analysis: If AI adoption leads to a decline in entry-level IT jobs, India could face a significant disruption in its employment landscape. Entry-level roles traditionally serve as the foundation for skill development and career progression. Their reduction could hinder the creation of a skilled workforce pipeline.

Policy and Institutional Response: India must prioritize large-scale upskilling initiatives that focus not only on coding but also on critical thinking, problem-solving, and AI collaboration. Educational reforms are crucial. For instance, the introduction of AI and computational thinking in CBSE curricula is a step in the right direction. Higher education institutions must also integrate AI-focused courses and practical training modules.

Industry and Individual Strategies: At the industry level, companies should invest in continuous learning programs for employees. Individuals, on the other hand, must adopt a mindset of lifelong learning and adaptability. For example, a software engineer could transition into roles involving AI model supervision or prompt engineering. Such adaptive strategies can help India transform this challenge into an opportunity for building a future-ready workforce.