Artificial intelligence is spreading faster than any earlier technology, and Asia and the Pacific is where its effects will be tested most sharply. The region holds some of the world’s AI leaders alongside countries still struggling with reliable electricity and internet. The Next Great Divergence, from the UNDP Regional Bureau for Asia and the Pacific, examines whether AI will close those gaps or deepen them.

1Why the report asks the question

The report starts from history. The Industrial Revolution opened a Great Divergence in income, health and education between early industrializers and everyone else. Later advances in medicine, agriculture and trade drove a Great Convergence that lifted hundreds of millions of people out of poverty, but inequality inside countries kept rising.

The authors argue AI belongs in the same story. It is a general-purpose technology, like electricity or the internet, that can write, translate, read images and support decisions. The scale of its economic gains is uncertain, but analysts broadly agree it will grow output rather than shrink it. The open question is distribution.

Asia and the Pacific is the natural testing ground. In 2024, Singapore’s income per person was nearly 200 times Afghanistan’s, and the gaps inside countries are also wide. The report uses the IMF’s AI Preparedness Index to show readiness ranging from world-leading to among the lowest anywhere.

2People: capabilities, security and the planet

The report finds AI can expand what people are able to do. AI tutors can work in local languages and ease teachers’ workloads, and a mathematics tutoring pilot in Bhutan is expanding. AI that reads chest X-rays is being piloted for tuberculosis screening in India and Bhutan. In northeast India, deep-learning flood forecasting reportedly raised accuracy from about 38 percent to 80 percent, with warnings up to 96 hours ahead.

The same tools can exclude. People missing from data are invisible to AI, and the report points to women, rural and indigenous communities, older people and the poor. Women in South Asia are up to 40 percent less likely than men to own a smartphone. Credit models trained mostly on urban male borrowers can mislabel women entrepreneurs and rural farmers as high risk. Children face privacy and well-being risks, and AI’s energy, water and mineral demands fall hardest on places with the least capacity to manage them.

3Economy: growth, jobs and livelihoods

Estimates of AI’s productivity gains vary widely, from roughly 0.4 to nearly 3 percentage points a year, and the report notes that past general-purpose technologies took years or decades to show up in the statistics. For developing economies that fully adopt AI, it cites estimates of two or more extra percentage points of annual GDP growth over a ten-year horizon.

Gains are expected first in frontier economies such as China, Singapore, Japan and the Republic of Korea, and inside countries in cities, large firms and skilled workers. Disruption is expected to land more heavily on routine and entry-level work. Female employment is nearly twice as likely as male employment to be highly exposed (4.7 versus 2.4 percent of jobs, per ILO research), and young workers entering the job market face fewer stepping-stone roles. The report also warns that growth models such as textiles, electronics assembly and business process outsourcing could be reshaped over a decade.

The gig economy shows both sides. A freelancer in Jakarta can reach global clients, but AI tools that let one worker do the work of several can squeeze rates and raise insecurity.

4Governance: strengthening or eroding trust

Used well, AI helps governments listen and deliver. Singapore’s OneService assistant has halved resolution times, and Bangkok’s Traffy Fondue platform logged nearly 600,000 resident reports in two years. Digital twins let agencies test policies before acting.

The risks center on trust. Opaque “black box” systems are hard to contest, biased data can lock in exclusion, and heavy reliance on automated scores can leave vulnerable households without recourse. Privacy and security stakes rise when large linked registries and cloud platforms are involved, particularly where laws and oversight are weak. Countries with strong digital foundations capture the governance dividend first, while others depend on foreign models and platforms they cannot audit or adapt.

5Seven principles and two sets of levers

The policy agenda is built on seven principles grouped under three pillars, with a stated aim of “leave no mind behind”.

The report says “hard” and “soft” levers must move together. Hardware without skills wastes investment, and skills without access cannot scale.

6A roadmap by time and by starting point

The report recommends sequencing actions over time, starting with “no-regrets” steps:

1. Immediate (0–12 months): extend electricity, internet and devices; set up an AI inclusion task force and an exclusion dashboard; start basic AI literacy and ethics training; update procurement guidelines; and address data sovereignty in law.
2. Medium term (1–2 years): invest in computing, local data centers and the energy to run them; set up data governance bodies; build inclusion metrics into digital projects and national plans; and pool costs regionally where needed.
3. Longer term (3–5 years): achieve measurable gains on inclusion metrics, refine regulation from evidence, scale proven pilots nationwide and secure financing that outlasts projects.

It also tailors priorities to each country’s capacity:

To keep policy honest, the report urges measuring outcomes rather than rollouts: failure rates of new systems (fingerprint mismatches in one Indian state’s early Aadhaar rollout reached as high as 49 percent), appeal and overturn rates for AI-influenced decisions, access and affordability, and usage, all broken down by gender, location, age and disability. It closes with playbooks for eight sectors: health, education, finance, agriculture and fisheries, urban and transport, biodiversity and climate, public services, and security and justice.

7Where Lao PDR appears in the report

Lao PDR is listed among the lower-capacity settings. The report notes that Cambodia and Lao PDR still have large numbers of young people without foundational literacy and numeracy, and that limited affordable internet and weak rural connectivity make it harder to extend the national ID system to remote areas. It describes Lao PDR’s National Digital Economy Development Strategy (2021–2030) as an early-stage effort with a focus on AI education and infrastructure.

For lower-capacity countries, the report suggests prioritizing affordable nationwide connectivity, shared regional data facilities, minimum digital ID and low-cost payments with strong privacy defaults, offline-capable AI for health, farming and early warning, a “starter-pack” data-protection law, and simple dashboards that track inclusion. It also encourages training local developers to fine-tune open-source models for languages such as Khmer and Lao.

8What it could mean for Lao businesses

The report is written for policymakers. The points below are ALBA’s reading of how its findings may apply to Lao companies and are not recommendations from UNDP.

9The bottom line

The report’s message is that AI can make rural clinics, classrooms, small firms and public services work better, but only if connectivity, skills, safeguards and participation are built alongside it. Where those foundations are thin, the same technology risks concentrating gains among those already ahead. The authors say the measure of success is not how smart machines become, but how much more empowered people are because of them.

Sources

1. UNDP Regional Bureau for Asia and the Pacific (2025), The Next Great Divergence. Primary source for all findings and figures in this summary; download link below.
2. UNDP, Human Development Report 2025: the global report this regional study builds on.
3. UNDP, Asia-Pacific Human Development Report 2024: Making Our Future: governance framework and human security context.
4. IMF (2025), The Global Impact of AI: Mind the Gap: macroeconomic estimates and the AI Preparedness Index cited in the report.
5. ILO (2025), Generative AI and Jobs: A Refined Global Index of Occupational Exposure: job-exposure analysis, including the gender gap.
6. GSMA (2024), The Mobile Gender Gap Report 2024: smartphone ownership gap for women in South Asia.

Figures are reported as presented in the UNDP report, which cites the underlying studies and datasets. Please check the original report before relying on any specific number.

Preview the report on this page

If the preview does not load, open the PDF in a new tab.