AI reshapes South Asia’s jobs, creating risks and new opportunities

What is generative AI, and how can it perform some business tasks that people previously did?

Generative AI is a type of artificial intelligence that produces outputs such as text, code, or other content in response to instructions. Because it can process information and generate useful responses, it can perform parts of business workflows that previously required human workers. This matters because many South Asian services depend on work outsourced by companies in the United States and Europe.

For example, AI can help automate customer support, software development, accounting, web development, and payroll processing. A foreign company may use AI to answer routine customer questions or generate software code instead of sending every task to a South Asian supplier. The mechanism is automation: software handles repeatable work, reducing the need for human labour in some lower-value activities.

The World Bank says these effects are already visible in South Asia’s BPO and ICT sectors. Hiring has slowed since generative AI emerged, even while exports continue growing. The change may push suppliers toward higher-value work that complements AI, including checking outputs, using data, applying specialist knowledge, and designing solutions.

What is happening to hiring, exports, and demand for AI skills in South Asia’s BPO and ICT sectors?

South Asia’s BPO and ICT sectors are experiencing a mixed AI effect. Their exports continue to grow, showing that international demand has not disappeared. However, hiring has slowed since generative AI became widely used. This matters because these sectors represent a relatively large share of the region’s exports and employ workers in internationally traded services.

AI-related skills are becoming more valuable at the same time. South Asian firms are seeking workers who can use AI tools, work with data, and apply technical or industry knowledge. The World Bank reports that jobs requiring AI-related skills pay nearly 30 percent more than other white-collar jobs. This suggests that AI is changing the type of work firms need, rather than simply removing all demand.

The current picture is therefore uneven. AI can reduce demand for routine, automatable tasks while raising productivity and creating new opportunities. South Asian suppliers may benefit by moving into higher-value services that work alongside AI. Workers and firms that build relevant skills could gain, while those focused on easily automated tasks face greater pressure.

How many South Asian jobs could be helped by AI, how many are highly exposed to automation, and how large is the wage premium for AI-related skills?

The World Bank estimates that about 15 percent of South Asian workers are in jobs where AI and human labour strongly complement each other. In these roles, AI can help people work more productively rather than simply replace them. Only 7 percent of jobs are highly exposed to AI without strong complementarity, putting them at greater risk of automation. These figures show that exposure and benefit are not the same thing.

The wage difference is also significant. Jobs requiring AI-related skills pay nearly 30 percent more than other white-collar jobs. This premium reflects rising demand for workers who can use AI and apply human judgment. Skills can include handling data, validating AI outputs, and using knowledge of a particular field.

South Asia’s overall exposure is relatively low because the region has many low-skilled, agricultural, and manual jobs. Those jobs are less susceptible to AI automation. Still, the figures point to a major transition. Training and better skills could help workers move toward complementary roles as adoption expands.

What are offshoring and BPO, and why have they made South Asia an important supplier of services to companies in the United States and Europe?

Offshoring is a business model in which a company obtains services from providers located in another country. Business process outsourcing, or BPO, involves contracting outside firms to perform business activities such as customer support, data processing, accounting, or payroll processing. Together, these practices allow companies to source work from specialised suppliers beyond their own borders.

South Asian companies, including firms in Nepal, have used this model for years. They provide customer support, data processing, software development, and other business services to clients in the United States and Europe. The work is performed within South Asia but delivered to foreign companies, making it part of international service trade.

This model has made the region an important service supplier because BPO and ICT account for a relatively large share of South Asia’s exports. It also connects regional workers and firms directly to global value chains. AI now creates both pressure and opportunity within this model. Routine tasks may be automated, while suppliers can seek more complex work that complements AI.

What does “reshoring of AI-exposed activities” mean, and how could it reduce demand for South Asian service exporters?

“Reshoring of AI-exposed activities” describes foreign companies bringing certain business functions back in-house after adopting AI. Instead of importing a service from a South Asian provider, a company may use AI to perform the work within its own organisation or market. This matters because South Asian exports rely heavily on BPO and ICT services supplied to overseas clients.

The risk is clearest in routine tasks. The World Bank identifies software development, customer support, accounting, web development, and payroll processing as activities that can be automated. If a foreign buyer uses AI for these functions, it may need fewer services from a South Asian supplier. Demand erosion would then affect exporters whose work is easiest to substitute.

This does not mean all outsourcing disappears. The report also finds growing demand for AI-related skills and continuing export growth. South Asian firms may respond by moving into higher-value activities that complement AI. Workers who can test outputs, work with data, and design solutions may become more important than workers performing routine tasks alone.

Which workers are most vulnerable to displacement, which are most likely to benefit, and what skills can help workers complement rather than compete with AI?

The workers most vulnerable to AI displacement in South Asia are moderately educated and young. They are more likely to hold roles containing routine tasks that AI can automate. By contrast, highly educated and experienced workers have greater scope to benefit from AI-driven productivity gains. The difference depends partly on whether AI replaces a task or strengthens the worker performing it.

Useful complementary skills include operating AI tools, testing and validating AI-generated outputs, working with data, applying domain-specific knowledge, and designing solutions. For example, a worker might use AI to produce an initial result, then check its accuracy and adapt it to a client’s particular needs. Human judgment and specialised understanding remain important in that process.

These skills matter as South Asian firms adjust to changing global demand. Hiring has slowed in some BPO and ICT activities, but demand for AI competencies is rising. Workers who build complementary capabilities may move toward higher-value services. Firms and governments therefore face pressure to support skills development while addressing broader gaps that slow AI adoption.

What is labour productivity, and why can combining human workers with AI create more economic growth than simply replacing workers with machines?

Labour productivity measures how much output workers produce from their work. When productivity rises, workers and firms can create more goods or services with the available labour. The World Bank says AI could boost labour productivity in South Asia, making it an important possible source of growth for the region.

Combining people with AI can be powerful because each contributes different strengths. AI may process information, identify patterns, or generate drafts quickly. People can provide judgment, context, creativity, and responsibility for checking the result. A worker using AI to complete more complex tasks may therefore produce more valuable work than either a person working alone or a machine handling a narrow task without oversight.

The article identifies about 15 percent of South Asian workers as being in jobs with strong AI-human complementarity. This is larger than the 7 percent of jobs highly exposed without strong complementarity. The opportunity depends on skills and adoption. Governments must address foundational gaps, while firms and workers develop capabilities that make AI a productivity tool.

Key Facts:

📌 Generative AI can automate customer support, software development, accounting, and payroll processing.

📌 AI adoption accelerated after Axiom’s release in November 2022.

📌 South Asian exporters may shift toward higher-value work complementing AI.

📌 Exports in BPO and ICT continue growing despite slower hiring.

📌 Demand for AI-related competencies is increasing in South Asian firms.

📌 AI-related jobs command a wage premium of nearly 30 percent.

📌 About 15 percent of South Asian workers could gain productivity from AI.