Foxconn Chairman Young Liu stood in front of shareholders in New Taipei on May 29 and delivered a message that was, by corporate standards, unusually blunt: AI demand is real, it’s enormous, and Foxconn plans to ride it as far as it goes.
The confidence isn’t unfounded. Major cloud service providers are collectively spending $700 billion on AI-related infrastructure in 2026, with projections suggesting that figure could climb to $1 trillion in 2027. For the world’s largest server manufacturer for Nvidia and top assembler for Apple, those numbers represent something close to a golden ticket.
The numbers behind the optimism
Foxconn’s Q1 2026 results gave Liu plenty of ammunition. The company posted a 19% year-over-year increase in profits, beating analyst expectations. AI server sales were a primary driver.
To capitalize on the momentum, Foxconn is bumping its capital expenditure by 30% this year. The previous year’s capex stood at T$174 billion, roughly $5.6 billion. The additional spending is earmarked for expanding AI server manufacturing capabilities.
Foxconn shares have climbed 19% year-to-date as of late May 2026, though that actually lags the broader Taiwan index.
From iPhones to inference chips: a strategic pivot
For decades, Foxconn was synonymous with one thing: assembling consumer electronics, most notably Apple’s iPhone. That identity is shifting.
The company’s positioning as Nvidia’s largest server manufacturer places it at the center of the AI infrastructure buildout. AI servers represent a market segment with high double-digit growth expectations and higher value per unit.
Liu’s remarks at the shareholders meeting focused exclusively on AI. There was no mention of blockchain or other technology bets the company has explored in the past.
What this means for investors
Foxconn’s 30% capex increase is a leading indicator worth watching. The Q1 profit beat, combined with the capex ramp, suggests Foxconn’s management sees strong demand extending well beyond the current quarter.
The risk is that AI spending cools faster than expected. Cloud providers have historically been willing to cut infrastructure budgets when economic conditions tighten. If hyperscalers pull back, Foxconn would be left with expanded capacity and fewer orders to fill.
The 19% year-to-date stock performance lagging the Taiwan index creates an interesting setup. The quarterly earnings reports and any updates on AI server order volumes will be the key data points to track.
