The reckoning is here. As Microsoft, Meta, Apple, Amazon, and Alphabet kick off earnings season this week, Wall Street isn't just looking for revenue beats - they're demanding answers on the biggest spending spree in tech history. The four hyperscalers are projected to pump over $470 billion into AI infrastructure in 2026, up from $350 billion last year, according to FactSet analyst estimates. But after OpenAI's commitments hit $1.4 trillion and bubble warnings intensified in Q4, investors want proof these data center bets will actually pay off.
If 2025 was Wall Street's crash course in AI infrastructure economics, 2026 is the final exam. And Microsoft, Meta, Amazon, and Alphabet are walking into earnings calls with a collective $470 billion spending plan that has investors equal parts excited and terrified.
The stakes couldn't be higher. After Meta's stock tanked 15% in October when the company raised its capex guidance to $72 billion - its worst day since 2022 - the message from Wall Street was clear: show us the money, or at least show us the path to profitability. The social media giant faces unique scrutiny because unlike its cloud-rich peers, it doesn't have an AWS or Azure to justify the spending. CEO Mark Zuckerberg tried to calm nerves by insisting "we're seeing the returns in the core business," but analysts at Deutsche Bank warn that "investor fears around the potential impact to earnings from the projected spend" could overshadow optimism.
The spending surge isn't happening in a vacuum. OpenAI's commitments reached $1.4 trillion in Q4, forcing the ChatGPT maker to ink massive deals with Nvidia, Broadcom, Oracle, Amazon, and Google just to keep the lights on. The entire industry is now riding on OpenAI's ability to keep raising cash - and that's making investors nervous about an AI bubble that could make the dot-com era look quaint.
Microsoft kicks off earnings on Wednesday facing tough questions about cost control. The company shocked analysts in October when CFO Amy Hood reversed course and said capex growth in 2026 would actually accelerate after previously signaling a slowdown. FactSet analysts now expect fiscal 2026 capex to hit $99 billion, with Q2 alone reaching $36.25 billion - up 60% year over year. The Visible Alpha consensus shows operating margins narrowing to 67%, the tightest in three years.
The silver lining for Microsoft? It's diversifying beyond OpenAI. The company announced a $5 billion investment in Anthropic in November, with the Claude maker committing to $30 billion in Azure compute capacity. Still, adoption of Microsoft 365 Copilot remains sluggish. KeyBanc analysts noted that "over half of organizations are licensing only up to 10% of the M365 user base" for the AI add-on that was supposed to drive software revenue growth.
Amazon is placing the biggest bet of all. Analysts project the e-commerce giant will spend over $146 billion in 2026, up 17% from $125 billion. The company's strategy got a major boost with its $38 billion AWS deal with OpenAI in November, its first contract with the ChatGPT maker. CNBC reported in December that Amazon is also in talks about a potential $10 billion investment in OpenAI, even as it continues backing rival Anthropic, which just raised $10 billion at a $350 billion valuation.
CEO Andy Jassy told investors last quarter that AWS was "gaining momentum," especially from AI workloads, but the cloud unit is losing ground to Microsoft's Azure, which has posted faster growth rates. Amazon reports earnings next week and will need to prove it can maintain its cloud infrastructure lead while juggling investments in multiple AI horses.
Alphabet enters earnings with serious momentum. Last year marked the stock's best performance since 2009 as investors regained confidence in its AI strategy after OpenAI's ChatGPT initially spooked the market. The company lifted its 2025 capex forecast in October to a range of $91 billion to $93 billion and warned of a "significant increase" in 2026. Analysts expect over $115 billion this year.
Google's dealmaking has been aggressive. It signed a multibillion-dollar deal with Anthropic in October that provides up to 1 million of its tensor processing units, and just landed a major win with Apple to power Siri's massive overhaul using Gemini models. Bank of America analysts said the Apple deal could drive a new iPhone upgrade cycle, though the financial terms remain unclear.
The wild card is whether OpenAI testing ads on ChatGPT in the U.S. will cannibalize Google's core search business - the very thing that funds this spending spree in the first place.
Apple occupies a different position in this arms race. The company isn't building massive data center infrastructure but has faced criticism for falling behind on AI tools. Its Siri revamp was delayed into 2026 after personalization features took longer than expected, and the company briefly disabled AI notification summaries for news in January after displaying inaccurate information.
CEO Tim Cook told CNBC's Steve Kovach in October the company was expecting the "best ever" December quarter with iPhone 17 reception "off the chart" after the device received positive reviews in September. Analysts are forecasting 10% to 12% revenue growth and double-digit iPhone growth, but investors want clarity on whether the Google Gemini partnership signals a strategic shift or an admission that Apple can't build competitive AI in-house.
Then there's Tesla, which reports Wednesday alongside Microsoft and Meta. CEO Elon Musk has spent years selling a "sustainable abundance" future where humanoid robots do every job, but investors want answers about the core auto business after deliveries fell 8.6% in 2025 to 1.64 million vehicles. Last quarter, shares slumped when Musk talked up Optimus and Robotaxi but avoided questions about fundamentals.
The company's energy unit grew last year, including sales supporting Musk's xAI, and investors are watching whether Tesla's board will invest in the OpenAI competitor. FactSet analysts project capex growing to $11 billion in 2026 from $9.5 billion, driven by new chip production with Samsung and TSMC announced at November's shareholder meeting.
The common thread across all these earnings calls? Executives have spent months insisting they can't build fast enough to meet AI demand. Now they need to prove those data centers, chips, and partnerships will generate returns before investor patience runs out. In October, Alphabet, Amazon, and Meta all raised spending guidance, while Microsoft's finance chief signaled more growth ahead. The market's response was brutal, particularly for Meta.
But the megacaps don't have the luxury of staying private like OpenAI or Anthropic. They need to show quarterly progress, maintain margins, and keep shareholders happy while placing trillion-dollar bets on an AI future that's still largely theoretical for most enterprises. Goldman Sachs projects Meta alone could hit $125 billion in capex this year, climbing to $144 billion in 2027.
This earnings season will define whether Big Tech's AI spending spree is visionary or reckless. With over $470 billion on the line in 2026 alone, investors aren't asking if these companies should build AI infrastructure - they're demanding proof it'll generate returns before the money runs out. Meta needs to explain how digital advertising justifies $110 billion in capex. Microsoft must show margin expansion despite record spending. Amazon has to defend its position against Azure's faster growth. And Alphabet needs to prove its search business won't get cannibalized by the very AI tools it's funding. The executives who spent 2025 saying they couldn't build fast enough now face a tougher question: when does fast enough become profitable enough?
