Energy markets were mixed on Tuesday, but the standout mover was GE Vernova, whose shares dropped nearly 7% after SoftBank disclosed a $3.73 billion loss tied to a subsidiary that buys power equipment for US artificial intelligence data-center projects.
The Japanese tech investment group said the hit came from a unit that sits in the procurement chain for data-center power gear. That unit had been placing orders for turbines, transformers, and other grid equipment to support the rapid buildout of AI infrastructure in the United States.
For investors, the news is a reminder that the AI boom's physical backbone—the data centers that train and run large language models—depends on a vast and complex supply chain. When one link in that chain stumbles, the effects ripple outward to suppliers like GE Vernova, which makes the heavy electrical equipment those facilities need.
Why the AI data center story matters for energy suppliers
Over the past year, the AI buildout has been a major driver of demand for power equipment. Data centers consume enormous amounts of electricity, and utilities and tech companies have been scrambling to secure turbines, switchgear, and other components to keep the lights on. That has given companies like GE Vernova a steady pipeline of orders, a contrast to the boom-and-bust cycles typical of many other energy projects.
But Tuesday's disclosure suggests that pipeline isn't without risk. SoftBank's loss indicates that the costs of procuring this equipment can be volatile, and that projects may face delays or cancellations. When a major player like SoftBank takes a hit, it can signal that the pace of data-center construction may not be as smooth as hoped.
GE Vernova, which was spun off from General Electric in 2024, has become a bellwether for the AI power trade. Its stock had rallied sharply on expectations of surging demand from data centers. Tuesday's drop shows how sensitive that trade is to any sign of weakness in the AI infrastructure spending cycle.
What this means for investors
For everyday investors, the key takeaway is that the AI story isn't just about chipmakers and software companies. It's also about the physical infrastructure—power generation, grid equipment, and cooling systems—that makes AI possible. Companies in that supply chain can be just as volatile as tech stocks, and their fortunes are tied to the pace of data-center construction.
The SoftBank loss also highlights the financial risks that tech conglomerates are taking on as they pour money into AI. SoftBank has been one of the most aggressive investors in AI, and its recent swing to a $7 billion loss underscores how costly this bet can be. The company's struggles could have broader implications for the sector, as other firms may become more cautious about their own AI spending.
For those holding energy or industrial stocks, the news is a cautionary tale. The AI-driven demand for power equipment is real, but it's not immune to setbacks. Investors should be prepared for volatility as companies navigate the complexities of building out the grid.
Meanwhile, the broader energy market was mixed: oil slipped while natural gas rose. That divergence reflects the different forces at play—oil is more tied to global demand and geopolitical news, while gas is increasingly linked to power generation, including the electricity-hungry data centers.
As the AI buildout continues, investors will be watching for signs of strain in the supply chain. Bond investors have already started pulling back from AI data-center debt, a sign that some are worried about the sector's ability to generate returns. If more companies like SoftBank report losses, that caution could spread.
For now, the immediate focus is on GE Vernova and other suppliers. Their earnings reports and guidance will be closely scrutinized for clues about the health of the AI power pipeline. Venture Global's bet on AI data centers as a growth engine is another sign of how deeply the sector is intertwined with the AI narrative.
In the end, Tuesday's news is a reminder that the AI revolution is not just a digital phenomenon—it's an industrial one. And like any industrial boom, it comes with fits and starts.
