Shares of major US technology companies rallied on Wednesday after Democrats secured control of both chambers of Congress in the midterm elections, easing uncertainty over the political outlook.
However, the results also raised concerns about the future of artificial intelligence infrastructure development, as voters rejected restrictions on data centers in several local ballot measures and Democratic candidates advocating tighter oversight won key gubernatorial races.
Amazon surged 8%, Google gained more than 6%, Microsoft advanced 4%, Meta rose 3%, while Apple climbed nearly 7% during Wednesday’s trading session. The gains suggested investors were responding positively to the prospect of a divided government, which could limit the scope of major policy changes and reduce uncertainty around corporate taxation, regulation and economic policy.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
David Kelly, chief global strategist at JPMorgan Asset Management, said the election results had removed an important source of uncertainty for financial markets.
Speaking on CNBC’s Squawk on the Street, Kelly said divided government could produce “moderation in growth, moderation in inflation, less uncertainty, and I think that’s all helping the stock market move up today.”
The rally was not uniform across the technology sector. Companies closely associated with President Donald Trump’s adviser Elon Musk and the expansion of AI data centers came under pressure as investors assessed the implications of the election results for infrastructure development.
Tesla fell 2%, while SpaceX declined 3%. Nvidia, whose processors power many AI data center projects, slipped just over 1%. Smaller AI cloud infrastructure providers suffered sharper losses, with CoreWeave dropping 10%.
The divergence points to a potential change in how investors assess the AI infrastructure market. While large technology companies have the financial resources, established cloud businesses and operating capabilities to manage regulatory challenges, smaller providers could face higher costs and greater uncertainty if new data center projects encounter political or regulatory obstacles.
A major concern for investors emerged from local ballot measures concerning data center development. Of the 41 measures considered on Tuesday to restrict or facilitate AI data centers, voters sided against data centers in 40, according to the report.
A CNBC Voter Poll also found that 56% of voters considered data centers a crucial or major factor in their voting decisions.
The results suggest that public opposition to large computing facilities has become a significant political issue. Communities have raised concerns about electricity demand, potential increases in power costs, water consumption and the pressure that data centers could place on local infrastructure.
The election also produced victories for Democratic gubernatorial candidates in Texas, Georgia and Ohio who had proposed moratoriums on data center development. Those outcomes have raised concerns that new projects could face additional regulatory scrutiny, delays or outright restrictions in states that have become important locations for technology infrastructure investment.
Macquarie analyst Paul Golding said several factors affecting data center construction fall under state-level oversight.
“Recall that backup power emissions, zoning, water, interconnect/grid, and tax considerations are often overseen by state regulators,” Golding wrote in a note on Wednesday.
The regulatory exposure is significant because building an AI data center involves more than securing funding and purchasing advanced processors. Developers must obtain land-use approvals, secure electricity connections, arrange water supplies and comply with environmental and local planning requirements. Restrictions in any of these areas can delay construction, increase costs or make a proposed site commercially unviable.
For technology companies committing billions of dollars to AI infrastructure, delays can affect the timing of capacity deployment and the expected returns on investment. The political environment is therefore becoming another variable in decisions about where to build and how quickly new facilities can become operational.
However, the election results do not automatically translate into a nationwide halt to data center construction. The ballot measures and gubernatorial victories point to potential restrictions in particular jurisdictions, while the broader demand for computing capacity remains intact.
Hyperscalers Gain An Advantage Over Smaller AI Cloud Providers
The market reaction suggests that investors see large cloud providers as better positioned to absorb regulatory and infrastructure pressures.
Hyperscalers such as Amazon, Microsoft, Google and Meta combine substantial capital expenditure with established cloud businesses and strong free cash flow. They also have experience developing and operating large data centers, giving them greater flexibility to change locations, negotiate infrastructure agreements and finance projects through periods of uncertainty.
Needham analyst Laura Martin argued that higher regulatory barriers could strengthen the position of established operators by making it more difficult for new competitors to enter the market.
“We believe higher regulatory barriers entrench incumbents, and increase risks (and raise required ROICs) for new entrants,” Martin wrote. “Hyperscalers are AI incumbents.”
That is considered necessary for neocloud companies, which provide specialized cloud computing capacity, often using high-performance graphics processing units to support AI training and inference. Unlike major cloud providers, many of these businesses have smaller balance sheets and rely more heavily on financing, infrastructure leases and customer commitments to expand.
CoreWeave’s 10% decline on Wednesday illustrated the pressure facing this segment. Baird analysts described the election outcome as a “sentiment headwind” for neocloud providers.
Golding said he remained positive on AI infrastructure investment but expected well-capitalized operators to benefit most if development slows in some states.
“Neoclouds may pay more to lease powered shells for GPUaaS commitments (and pass on costs), while hyperscaler/AI lab capacity needs could intensify competition for even more limited supply,” he wrote.
His assessment points to a possible increase in the cost of securing data center capacity. If regulatory restrictions limit the number of suitable sites, companies that lease powered facilities could face higher prices. Those additional expenses may eventually be passed on to customers, potentially weakening the competitiveness of smaller AI cloud providers.
Chris Miglino, chief executive of Axe Compute, cautioned against interpreting the election results as evidence of a nationwide slowdown in construction. He said companies were still raising capital, enterprises continued to spend on AI, and demand for additional computing capacity remained strong.
He argued that tighter regulation in some regions could lead investors and customers to pay closer attention to which providers can deliver capacity reliably, including companies offering AI inference services tailored to specific industries.
“If you’ve already secured power, have the financing in place and understand how to deploy and operate compute infrastructure efficiently, you’re in a much stronger position than a company that’s still trying to figure those things out,” Miglino said.
That assessment captures the emerging divide in the AI infrastructure market. The ability to raise capital and purchase processors remains important, but access to electricity, approved development sites, and operational expertise will increasingly determine which companies can bring computing capacity online.