Microsoft and Alphabet have signaled a more measured near-term pace for artificial intelligence infrastructure spending, with both companies reporting capital expenditure figures below analysts’ expectations as delivery delays and financing pressures expose vulnerabilities in the supply chain supporting the AI boom.
Microsoft lowered its capital expenditure forecast for the year to $170 billion from $175 billion, while Alphabet maintained its guidance at between $195 billion and $205 billion. Although neither company indicated that its broader AI investment programme was approaching a peak, the figures raised questions about the timing of infrastructure deployment and the ability of specialist cloud providers to deliver computing capacity when technology giants need it.
Shares of neocloud companies CoreWeave, Nebius and IREN fell following the earnings reports, as did Oracle, one of the largest investors in AI infrastructure. The market reaction suggests that investors are increasingly sensitive not only to how much hyperscalers plan to spend, but also to when that spending translates into operational capacity and revenue for the companies supplying it.
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.
Microsoft Chief Financial Officer Amy Hood said on the company’s earnings call on Wednesday that delays among third-party providers of graphics processing units would push some available computing capacity into the first half of 2027. Hood did not identify the providers, but Microsoft relies on CoreWeave, IREN and Nebius, among others, to supplement its own infrastructure.
Alphabet finance chief Anat Ashkenazi told analysts that external computing capacity “came online only at the very end of Q3.” The company’s shares fell about 10% in extended trading after it missed earnings and revenue estimates and reported negative free cash flow of more than $9 billion.
The disclosures point to a potential mismatch between the AI industry’s enormous investment commitments and the pace at which new computing resources become available. Delayed deliveries can defer revenue for infrastructure providers while leaving customers competing for scarce capacity, complicating forecasts for both sides of the market.
The risks are significant for neocloud providers, which specialize in supplying computing infrastructure, including GPU capacity, to AI developers and major technology companies. Their growth depends on securing expensive hardware, building data centers, and obtaining sufficient financing to support expansion before all the expected revenue arrives.
These companies have become important partners for hyperscalers seeking to expand computing capacity beyond their own facilities. However, they generally require billions of dollars in external funding and lack investment-grade credit ratings, leaving them exposed to changes in borrowing costs and lenders’ willingness to finance additional projects.
Rising interest rates and mounting concerns about debt burdens are making that funding model more difficult. If lenders become less willing to provide capital, neocloud operators could face pressure to slow construction, renegotiate commitments, or seek additional equity financing. Any resulting delays could then affect the technology companies depending on them for computing resources.
The problem is becoming more consequential as leading AI developers compete to secure infrastructure for demanding models and services. Anthropic relies on computing capacity from Amazon, Google and Microsoft, alongside providers including SpaceX, Lambda and Nscale. Nscale filed to go public in September but had not updated its prospectus in weeks, according to the report.
Anthropic’s Claude assistant has experienced periods of unavailability this year as demand increased. Rival OpenAI, which has said an initial public offering in 2027 is likely, also needs additional capacity to support its expanding operations.
For these companies, infrastructure availability is not simply a matter of long-term capital planning. Insufficient computing resources can constrain product deployment and service reliability, potentially affecting customer acquisition and revenue growth. But securing capacity too aggressively can leave them exposed if hardware deliveries slip, financing becomes more expensive, or demand fails to meet expectations.
The pressure is extending beyond individual companies. Gartner analysts told CNBC last week that the firm expects 80% of neocloud companies to be absorbed by competitors or cease operating entirely by 2030. The forecast underscores concerns about the sector’s ability to sustain its rapid expansion when many providers depend on continued access to outside capital.
Firmus, an Australian neocloud backed by Nvidia, withdrew plans for an initial public offering earlier this month despite reporting substantial usage commitments from Meta and OpenAI. The decision points to the line between attracting major customers and securing the financing needed to build infrastructure at scale. Commitments can support a business case, but they do not eliminate execution, funding, or cash-flow risks.
AI Investment Remains on Course to Accelerate
The spending adjustments at Microsoft and Alphabet do not amount to a retreat from the AI infrastructure race. Instead, they suggest that the pace of expenditure can be affected by practical constraints, including component costs, delivery schedules and the availability of external computing resources.
Hood said Microsoft’s capital spending could accelerate in the first half of 2027 because of higher component costs and delayed deliveries. The postponement of some capacity may therefore shift expenditure and deployment across reporting periods rather than eliminate the underlying investment.
Across the largest cloud providers, capital expenditure is expected to exceed $1 trillion next year, with some forecasts anticipating further increases thereafter. The scale of those projections highlights the importance of financing conditions and infrastructure execution to the technology sector’s earnings outlook.
Meta’s latest results offer another indication of the financial trade-offs. The company said last week that it expects full-year capital expenditure of $135 billion to $145 billion, raising the lower end of its previous range from $130 billion. Its shares also declined after the company reported negative free cash flow.
Alphabet faced a similar market response, although its disappointing results in the core search business also contributed to the sell-off. The combination of weaker-than-expected operating performance and substantial cash expenditure made it harder for investors to assess when AI investment might translate into stronger financial returns.
Microsoft, by contrast, received a more favorable reaction. The company beat earnings and revenue estimates, supported by 48% sales growth in its Azure cloud business. Investors appeared more willing to look past the near-term capital expenditure shortfall because of evidence that demand for its cloud services was translating into revenue growth.
“We view the quarter as further validation of Microsoft’s AI-led growth opportunity,” Cantor analysts led by Thomas Blakey wrote in a note to clients on Wednesday.
The analysts recommended buying Microsoft shares and raised their one-year price target to $630 from $608.
Amazon’s results, due after the market closes on Thursday, will provide another test of the industry’s spending outlook. Investors will be watching not just its capital expenditure plans, but also evidence of demand, cloud growth and the financial returns expected from its expanding infrastructure.
However, there is another concern. The AI industry’s ability to convert unprecedented spending into sustainable earnings without allowing financing costs and execution delays to undermine the business models supporting the expansion has come under scrutiny. Hyperscalers retain substantial resources to fund infrastructure, but their specialist suppliers face a more difficult balance between investment, debt, and cash generation.