AI exports power Asia’s surge, driving nearly half of 2025 economic growth: Nomura study

Rising demand for artificial intelligence has become a major driver of Asia’s economic growth since 2025, according to a study by Nomura economists.

The findings show that AI’s role in Asia’s growth has increased sharply. In 2024, AI contributed only 0.26 percentage points to Asia’s real GDP growth of 4.1%. However, in 2025, its contribution jumped to 2.01 percentage points. This means nearly half of Asia’s growth in 2025 was driven by AI.

AI-related net exports emerged as the biggest contributor

Over the past two years, AI-related net exports have been the biggest contributor to AI-driven growth. This has been especially visible in Taiwan, South Korea and Singapore, which are key suppliers in the global AI hardware and intermediate goods supply chain.

Thailand also benefited from higher AI-related exports. In contrast, in India and Indonesia, overall growth remains largely driven by non-AI demand. In these countries, AI’s contribution is either very small or negative.

In several economies, AI-related growth partly or fully offset weak performance in non-AI sectors. This helped some countries avoid an overall economic contraction. In contrast, in 2024, AI mainly provided a small boost rather than acting as a stabilising factor.

AI exports are slowly supporting capital expenditure

The report also notes that rising AI exports are slowly supporting capital expenditure. Investment in machinery and equipment is increasing. However, data centre construction is not yet a major growth driver, except in Malaysia and Singapore. This shows that these two countries are emerging as regional hubs for AI infrastructure.

Nomura expects AI’s importance in Asia to grow further, mainly through trade. It also expects data centres to play a bigger role in the coming years as AI adoption expands.

The economists said they will continue to update their estimates as more data becomes available. They believe this framework can help identify which Asian economies are most exposed to the AI cycle, both in terms of gains and risks.

The methodology 

The report says it is difficult to measure the exact impact of AI because official data does not have a separate category for AI-related activity. To address this gap, Nomura used an alternative method. It studied how AI demand affects three main areas of the economy: trade, investment and consumption.

Using the expenditure method of GDP calculation, the economists estimated AI’s contribution by measuring the year-on-year change in AI-related activity and comparing it with the previous year’s real GDP.