IMF warns of slowing Asia factory activity amid AI demand correction

Asia’s factories are still growing. They’re just growing with noticeably less enthusiasm than before.

Private surveys released on Monday, November 2, showed manufacturing activity across the region stayed in expansion territory in October. Growth cooled, however, in the economies most tied to the AI chip trade: Japan, South Korea and Taiwan. That pattern suggests the artificial intelligence boom may be losing some momentum as markets begin to question how profitable the sector really is.

The timing matters. The International Monetary Fund has already put a potential correction in AI-driven demand on its list of risks for the region. The October numbers are the first hard data that look a bit like that warning taking shape.

The numbers behind the slowdown

According to S&P Global’s October surveys, South Korea’s PMI came in at 51.3, its lowest reading in eight months. Japan registered 53.3, a seven-month low.

Taiwan, home to much of the world’s advanced chipmaking, posted 54.9. That’s still solid growth, but it’s down from 56.7 the prior month.

China was the odd one out, and not in a good way. Its official manufacturing PMI slipped to 49.3, meaning activity there actually contracted rather than merely slowing.

A war, an inflation problem, and a chip question

Some analysts have linked the trends to waning demand in semiconductor markets alongside inflated prices driven by persistent inflationary pressure.

What the IMF is forecasting

The IMF’s October 2026 Regional Economic Outlook for Asia and the Pacific projects regional growth will moderate to 4.8% in 2026, down from 5.0% in 2025. It then sees a further step down to 4.4% in 2027.

Advanced economies in the region are expected to see growth accelerate to 3.2% in 2026, with the AI sector doing much of the lifting. Emerging economies, by contrast, are forecast to moderate to 5.1%.

The IMF flagged three main risks in its outlook: a potential correction in AI-driven demand, ongoing geopolitical tensions affecting energy prices, and rising global yields.

The fund also modeled what a hypothetical sharp AI demand correction could do. Its estimate suggests Asia’s output could fall by around 1.7% below baseline, with the hit landing on technology supply chains.

South Korea runs the stress test

South Korea’s central bank has done its own math, and it isn’t exactly soothing. The Bank of Korea warns that a slowdown in AI-related capital spending could cut the country’s semiconductor exports by between $50 billion and $100 billion.

That kind of export loss could reduce South Korea’s 2027 GDP growth by 0.4 percentage points, according to the central bank.

The IMF has also looked at the labor side of the AI story. Its analysis of around 40 million job postings found employer demand for AI-exposed occupations has declined by nearly 10% since ChatGPT launched in November 2022.