Meta shares slide as AI spending weighs on cash flow and outlook

Traders Union
All News
Financial News
Meta shares slide on AI spending

Highlights

Meta forecasts current-quarter revenue of $61bn to $64bn, missing Wall Street's $63.1bn expectation at midpoint, with costs up 55% to $42bn.
Free cash flow plunges 91% year on year to $784mn as AI infrastructure spending accelerates; 2026 capex forecast rises to a $130bn–$145bn range.
Meta enters early talks to rent compute power to Anthropic in a deal worth up to $10bn, as it pushes to commercialize AI assets and counter rivals.

Quarter results and higher spending guidance

As reported by Financial Times, Meta forecasts current-quarter revenue of $61bn to $64bn, with the midpoint below Wall Street expectations of $63.1bn, while second-quarter sales come in only slightly above analyst estimates. Costs and expenses rise 55% to $42bn, and free cash flow for the three months to the end of June falls 91% year on year to $784mn from $8.5bn as AI infrastructure spending accelerates.

The company says second-quarter capital expenditure stands at about $31bn. Meta also narrows its 2026 capital expenditure forecast to $130bn to $145bn, from the $125bn to $145bn range given in April, and lifts the lower end of its 2026 total expense outlook to $165bn to $169bn, citing $2.4bn in legal proceeding charges recognized in the second quarter.

Net income declines 14% to $15.8bn, well below analysts' estimates of $18.5bn, hurt by one-off legal costs and severance payouts. Earlier this year, Zuckerberg cuts the workforce by about 10%.

AI agents strategy faces investor test

On a call with analysts, chief executive Mark Zuckerberg says Meta is at a point where its AI investments are accelerating every major part of the core business, including advertising and content recommendation systems. He presents a strategy centered on AI agents that can carry out tasks for users around the clock, arguing that personal agents for consumers can form the foundation of the next wave of products and revenue lines in the months and years ahead.

Zuckerberg also highlights business uses including coding tools, advanced customer service agents and what he describes as a potential business-in-a-box service on Meta. He says a significant portion of Meta's compute capacity is expected to support its own models and products, but adds that the company is also receiving offers to sell compute at a significant premium to its cost.

Meta is in early talks to rent computing power to Anthropic in a deal that could be worth as much as $10bn, moving into a market dominated by Amazon and Google. The company remains under pressure to show that its AI push can close gaps with OpenAI, Anthropic and Google after its Llama models fall behind last year, even as its new TBD Lab team under Alexandr Wang releases newer models such as Muse Spark and works on future versions including one codenamed Watermelon.

Mike Proulx, research director at Forrester, says Meta views AI infrastructure as a strategic asset, but the cost is arriving faster than the payoff. That leaves investors weighing whether the company's expanding AI projects amount to diversification or distraction as Meta also tests new monetization options such as subscriptions for advanced features and more complex AI tasks across its apps.

Our earlier article on Microsoft’s fiscal Q4 earnings recap covered an above-consensus revenue print and strong momentum in its cloud segment, with Azure growth accelerating and annual Azure revenue surpassing $100 billion. We also highlighted rising commercial backlog driven by customer commitments beyond large AI developers, alongside investor focus on generative AI competition and the company’s capacity and partnership-related risks.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.