The last few days saw tech giants endure an AI massacre by DeepSeek — but now that things have settled, will these companies survive the after-effects during their earnings? The Magnificent Seven tech companies are betting trillions on AI infrastructure, while OpenAI, SoftBank Group ($SFTBY), and Oracle ($ORCL) are leading the largest project, Stargate, which is set to deploy $500B inti AI investment over the next four years. But investors are growing impatient, wondering when these data center spend-a-thons will pay for themselves.

Show me the money: Jefferies has cautioned that DeepSeek’s efficient method “punctures some of the capex euphoria,” with Emarketer principal analyst Jeremy Goldman noting that the massive infrastructure investments of tech giants powerhouses “bloated, not visionary.” As a result, tech leaders are on the defensive to start their earnings season — and explain how they’ll respond to DeepSeek’s blindsiding advances. Unsurprisingly, MetaMETA

and MicrosoftMSFT

insist it’s full speed ahead. Despite spending a combined $37.4B on AI this quarter alone and beating overall earnings expectations on Wednesday, Microsoft and Meta couldn’t erase market concerns, assuring that AI spending will be a recurring theme this year.

• Microsoft’s cloud division missed expectations despite 157% YoY growth in AI services revenue, while the company doubled its capital expenditure to $22.6B — primarily for AI infrastructure.
• Despite Meta projecting $117B in expenses for 2025, up 22% from 2024, CEO Mark Zuckerberg remains undeterred, projecting up to $65B in AI-related investments for 2025.

Is AI’s ROI Just a Mirage?

The artificial intelligence gold rush has companies emptying their coffers — even selling furniture and laying off employees — to fund their AI ambitions. But whether there’s an audience large enough to cover their data center bills is still unknown. Tech firms have been pushing AI features onto individuals and bundling features in upgraded enterprise software — but with more companies pushing AI products, the return on investment is yielding mixed results.

• ServiceNow’sNOW
  AI bet hit turbulence after shifting to “pay-as-you-go” pricing for generative AI tools — delaying significant revenue generation and impacting its short-term sales outlook.
• Meanwhile, IBMIBM
  reaped the rewards of its AI investment, reporting $5B in bookings for its generative AI business, which includes sales and future sales in its software and consulting sections.

The long game: While AI enthusiasm remains high — with 98% of Fortune 1000 data leaders planning to increase AI spending in 2025 — the technology’s most significant impact isn’t where many expected.  GraniteShares founder Will Rhind cautions that AI’s payoff timeline needs patience, noting, “This process is going to need to take some time to play out.” Northern Trust’s Katie Nixon echoes this sentiment but remains bullish on the Magnificent Seven’s prospects, stating, “I wouldn’t bet against them.” In any case, Meta CEO Mark Zuckerberg says we’ll have a clearer idea of how the AI business is playing out at the end of the year — but the world will be watching closely.