Meta, Apple, Tesla, Microsoft AI spend in focus

If 2025 was the year Wall Street came to grips with massive artificial intelligence infrastructure spending from tech’s megacaps, 2026 looks to be more of the same.

But as the price tag for AI goes up, so do expectations for the returns on investment.

Earnings season kicks off this week for tech’s biggest names, with reports from Apple, Meta, Microsoft and Tesla. Next week features earnings announcements from Alphabet and Amazon.

It’s the first opportunity for industry leaders to clearly lay out their spending visions for the year as AI dealmaking accelerates and companies move from announcing new data centers to constructing them.

It’s also a chance for investors to hear how and when those projected build-outs are expected to turn profitable.

In total, the four so-called hyperscalers — Microsoft, Meta, Alphabet and Amazon — are expected to boost capital expenditures this year to over $470 billion from about $350 billion in 2025, according to analyst estimates compiled by FactSet.

As they address analysts, some CEOs are likely to find themselves in defense mode, justifying their investments after sentiment soured in some capital-intensive corners of the market late last year.

To date, executives have repeatedly said that they can’t build out fast enough to meet the insatiable demand for new models and services.

In October, Alphabet, Amazon and Meta all upped their spending guidance for 2025, and Microsoft’s finance chief said higher growth was on the horizon.

Meta’s stock had its worst day in three years after the company lifted its spending forecast, with investors concerned that the social media company is most at risk of racking up losses on its infrastructure because it doesn’t have anything resembling a cloud computing business.

Chatter of an inflating AI bubble picked up in the fourth quarter, as OpenAI’s commitments reached $1.4 trillion, meaning the ChatGPT maker needs to keep raising hefty amounts of cash to fund its plans. And those plans are increasingly tied to the fate of the rest of the tech industry.

OpenAI announced multibillion-dollar agreements with Nvidia, Broadcom, Oracle, Amazon and Google as it lessened its reliance on Microsoft, which long served as the company’s anchor partner and investor. A year ago, OpenAI and Microsoft ended an exclusive cloud agreement.

But unlike OpenAI or Anthropic, which remain private companies, the megacaps need to show that the aggressive dealmaking is supporting a grand plan, while also growing revenue and keeping investors happy.

Here’s what Wall Street is expecting as tech earnings season kicks into gear.

Microsoft

Microsoft has to show that it can control costs as it builds out data centers to meet AI demand and to support its Azure cloud unit.

The stock dropped in October after the company upped its spending guidance, and CFO Amy Hood said capex growth in 2026 would mark an increase from 2025, after previously saying that growth would slow.

Analysts polled by FactSet expect capex to rise to $98.8 billion this fiscal year, which ends in June, and jump again over the next two years. The Visible Alpha consensus for fiscal second-quarter capital expenditures and finance leases was $36.25 billion, up 60% year over year.

In October, the company called for its operating margin to be flat year over year, while analysts polled by Visible Alpha foresee the narrowest operating margin in three years, at 67%.

While OpenAI diversifies away from Microsoft, the same is true from the other side.

In November, Microsoft announced a strategic partnership with Anthropic that included a $5 billion investment in the Claude maker. Anthropic committed to buying $30 billion of Azure compute capacity.

The company’s big growth play remains in cloud infrastructure.

In its last earnings report, Microsoft called for 37% growth in revenue at constant currency from Azure infrastructure and other cloud services for the current period, slipping from 39% at constant currency in the September quarter.

Analysts from Evercore ISI said in report last week that, after attending a Microsoft AI Tour event in New York, they felt that Azure continued to enjoy a “healthy competitive position.”

One big question for Microsoft remains adoption of its enterprise AI services, notably the Microsoft 365 Copilot add-on, as it’s viewed as a source of revenue growth for the company’s software suite. KeyBanc analysts, in a note on Jan. 22, offered some reasons for concern.

“We heard from one partner that over half of organizations are licensing only up to 10% of the M365 user base, while just under 25% of these organizations are licensing Copilot for up to 25% of the user base,” wrote the analysts, who recommend buying the stock.

Meta

Meta gets almost all of its revenue from digital advertising, a reality that has left some investors puzzled as the company has ramped up investments in AI with no clear monetization story.

Furthermore, its costly AI strategy has shifted over the last few months following a failed launch of its latest Llama model. The big move was a $14.3 billion investment in Scale AI in June, which brought over CEO Alexandr Wang and other top talent to the company.

In its October earnings report, Meta lifted its 2025 guidance for capital expenditures to between $70 billion and $72 billion from a prior range of $66 billion to $72 billion. CEO Mark Zuckerberg insisted that Meta was investing in AI for a future big payoff.

“We’re seeing the returns in the core business that’s giving us a lot of confidence that we should be investing a lot more, and we want to make sure that we’re not underinvesting,” he said.

Analysts polled by FactSet are projecting nearly 57% growth in capital expenditures in 2026 to over $110 billion. Goldman Sachs sees that number going even higher, forecasting capex this year of $125 billion, going to $144 billion in 2027.

“Investor fears around the potential impact to earnings from the projected spend, as well as reduced financial flexibility from the elevated investments in the near-to-mid-term, could somewhat outweigh optimism around faster growth,” analysts at Deutsche Bank said.

Last month, CNBC reported that Meta was working on new frontier AI model known internally as Avocado.

Apple

Apple is fresh off a high-profile deal with Google to use its Gemini models for a massive Siri overhaul.

In the spring, the company had pushed off a revamp of its flagship Siri voice assistant after warning that certain personalization features would take longer than expected to deliver. Analysts at Bank of America said the agreement with Google could be a major driver for iPhone upgrades in the months ahead.

Apple has long been at risk of falling behind rivals such as OpenAI and Google on AI tools. And, while the company is growing its AI strategy, it’s doing so at a much slower pace. Apple hasn’t made a major AI announcement since the launch of Apple Intelligence in 2024, and even that rollout brought its own obstacles.

Last January, the company briefly disabled AI notification summaries for news after it displayed inaccurate facts.

Investors will be closely watching for signs of any shift in the company’s AI strategy or higher capital expenditure costs. They’re also monitoring for signs of an iPhone super-cycle, with analysts expecting a big quarter after the iPhone 17 launch in September received positive reviews.

Apple said in October that it expects 10% to 12% revenue growth in the current quarter and iPhone revenue growth in the double digits year over year.

At the time, CEO Tim Cook told CNBC’s Steve Kovach that the company was expecting the “best ever” December quarter in its history and that reception for the Phone 17 devices was “off the chart.”

Amazon

Amazon upped its capex forecast in October to $125 billion for 2026, from $118 billion, due to demand for its AI services. That was the highest spending forecast among the megacap companies.

Analysts are forecasting 17% growth in 2026 to $146 billion, according to FactSet.

Amazon has long been a leading provider of cloud infrastructure technology, but has also faced growing pressure from investors to explain its AI strategy and prove it can better compete against companies like OpenAI, Google and Anthropic.

In November, Amazon Web Services signed a $38 billion deal with OpenAI, its first contract with the ChatGPT maker. As part of the deal, OpenAI would run workloads on AWS infrastructure, using Nvidia’s graphics processing chips.

The following month, CNBC reported that the e-commerce giant was in talks over a potential $10 billion investment in OpenAI. Amazon has long backed OpenAI competitor and Claude maker Anthropic, which recently raised a $10 billion funding round at a $350 billion valuation.

While AWS leads the cloud infrastructure market, Microsoft’s Azure has been growing faster. Amazon CEO Andy Jassy said on the company’s third-quarter earnings call that the business was “gaining momentum,” especially from AI workloads.

Alphabet

Last year proved to be a big spending year for Alphabet, but it was also the best year for the stock since 2009 as investors gained confidence in its AI strategy.

Alphabet in October lifted its 2025 capex forecast to a range of $91 billion to $93 billion. Alphabet previously upped the range to $75 billion to $85 billion in July due to robust cloud products and services demand.

For 2026, analysts expect over $115 billion in spending from the search giant.

Over the last year, Google has inked deals with both OpenAI and Anthropic. In October, Anthropic and Google Cloud signed a multibillion-dollar deal that would bring over a gigawatt of AI compute capacity this year. And the company will need to add more capacity to accommodate as many as 1 million of its tensor processing units from the deal.

Alphabet is also fresh off a deal with Apple to use its Gemini model for the iPhone maker’s massive Siri overhaul. The deal, announced earlier this month, was another major boost of confidence in Google’s AI revival after OpenAI got off to a hot start with ChatGPT.

Google already shells out billions to Apple each year to be the default search engine on iPhones. Terms of the latest arrangement will be of interest to Wall Street.

Investors will also be watching for signs of ongoing search growth, and indications that AI hasn’t cannibalized the company’s core business. OpenAI said earlier this month that it would soon begin testing ads on ChatGPT in the U.S.

Tesla

The story for Tesla looks slightly different than it does for its peers.

For years, Elon Musk has sold investors on the vision of a “sustainable abundance” future, where robots outnumber people and do every job imaginable.

Investors more focused on the current year will want updated guidance on Tesla’s core automotive and energy sales. The company’s automotive deliveries fell 8.6% in 2025 to 1.64 million, from 1.79 million in 2024.

Meanwhile, Tesla’s energy unit, which sells battery energy storage systems for use in homes, businesses and massive utility-scale projects, grew last year.

Part of the company’s energy sales also supported Musk’s AI company, xAI, and investors will be watching to see if the carmaker’s board plans to invest in the OpenAI competitor.

Wall Street also wants to see that the company can show future growth and profit from its newer ventures, including its Robotaxi ride-hailing service launched in 2025 and its Optimus humanoid robots that have yet to go on sale.

Last quarter, Tesla shares slumped after Musk talked up the company’s Optimus and Robotaxi efforts but failed to confront questions about the fundamentals of the auto segment.

Investors will also be watching Tesla’s planned capex, especially for the chip technology that will underpin future autos and robotics.

During Tesla’s annual shareholder meeting in November, Musk said the company would be moving ahead with production of new chips with Samsung and Taiwan Semiconductor Manufacturing.

Analysts polled by FactSet expect capex to grow to $11 billion this year from a projected $9.5 billion in 2025.

— CNBC’s Jordan Novet, Jonathan Vanian, Lora Kolodny, Kif Leswing, Jennifer Elias and Annie Palmer contributed reporting