Is artificial intelligence a bubble? Does the demand for the technology justify the massive capital investment in AI data centers? While the answers to those questions still aren't clear, Taiwan Semiconductor (TSM +0.25%) Chairman and CEO C.C. Wei now appears convinced that AI is a big deal.
TSMC reported stellar results for the fourth quarter of 2025. Revenue rose 20.5% year over year, net income jumped 35%, and the company provided solid guidance for 2026 based on persistently strong demand. TSMC sees revenue growing by nearly 30% this year, with a 25% compound annual growth rate through 2029.
Image source: Getty Images.
A major driver behind that forecast is demand for AI accelerators. Through 2029, TSMC expects AI accelerator revenue to grow by at least 50% annually. Major players in the AI industry have reportedly been frustrated by TSMC's sluggish expansion, given the booming demand for AI chips and the chronic shortage of advanced semiconductor manufacturing capacity to produce them. TSMC tends to be conservative and risk-averse, but the company is now ramping up spending after becoming more comfortable with the future demand picture. TSMC isn't going all out, but it's increasing its capital spending to pursue the AI opportunity.
"AI is real"
For TSMC, worrying about an AI bubble makes a lot of sense. Building new leading-edge semiconductor foundries takes years and consumes many billions of dollars in capital. If TSMC is overly optimistic and overbuilds, its profitability could suffer tremendously.
"I'm also very nervous about it. You bet," said Wei in response to an analyst question about the trajectory of AI demand. Wei continued, saying that if the company wasn't careful with its capital spending, it would be a "big disaster" for TSMC.
While TSMC continues to take a conservative approach, Wei has spent the past few months talking to TSMC's customers and the customers of those customers. The aim of these discussions was to gauge whether AI demand was real. In other words, is AI actually helping these businesses?
Wei's conclusion is that "AI is real," calling it an "AI megatrend." TSMC's customers are chip designers like Nvidia and AMD, while its customers' customers are hyperscale cloud providers and other buyers of AI accelerators. The financial status of those hyperscalers also gave Wei come confidence. "They are very rich," Wei said.
Another data point is TSMC's own use of AI. Wei noted that the company is using AI to improve the productivity of its fabs, achieving 1% to 2% gains at essentially no cost. Wei didn't go into any details about those productivity gains.
Given the company's newfound confidence in the durability of AI demand, TSMC said it expects to spend between $52 billion and $56 billion in capital expenditures this year. For comparison, TSMC's capex in 2025 was around $40 billion. The company also expects capital spending to increase further over the next few years.
TSMC is also pulling forward the start of production at its second fab in Arizona to help meet demand. The company now expects high-volume manufacturing to begin in the second half of 2027, and it recently purchased a second plot of land to support future expansion.
Risks remain for TSMC
The semiconductor industry goes through cycles, but the current AI-driven cycle is a different animal. Wei noted that strong demand from the AI industry looks like it will go on for many years, calling it "endless."
While there's no question that demand for AI accelerators is extremely strong right now as hyperscalers plan and build massive AI data centers, TSMC is still taking some risk by ramping up its capital spending. If you go down the chain of customers to the next layer, to companies buying AI computing capacity from the hyperscalers, the demand picture is far more muddled.
AI companies like OpenAI, Anthropic, and xAI are burning cash at an incredible rate, partly to support massive user bases that don't pay the companies a dime. OpenAI reportedly has more than 800 million users for ChatGPT, but only a small percentage are on a paid plan. OpenAI reportedly plans to burn through $115 billion in cash through 2029, and xAI is reportedly incinerating nearly $1 billion in cash per month.
NYSE: TSM
Key Data Points
For TSMC, the sustainability of this spending is the biggest risk. Businesses and individuals are certainly using AI for real, productive purposes. However, it's not clear how much demand for AI computing capacity, and thus demand for AI accelerators, and thus demand for TSMC's manufacturing capacity, is being propped up by unsustainable spending from AI companies.
TMSC is still being fairly conservative despite the capital spending increase. The company may be leaving money on the table over the next few years, but its caution is ultimately a good thing for investors. If AI is a bubble and it pops, TSMC will get hit, but not as badly as it would if it had recklessly expanded production.
For at least the next few years, TSMC's revenue is likely to grow at a blistering rate thanks to AI. After that, it remains anyone's guess.
