Meta's Q3 2026 Results Show How AI Spending Is Reshaping Big Tech Earnings

Quarterly earnings season has become the main venue where investors learn exactly how much the largest technology companies are spending on artificial intelligence, and whether that spending is starting to pay off. Meta Platforms' third-quarter 2026 results, published on October 21, offer a clear recent example, pairing strong revenue growth with a sharply higher capital-expenditure outlook and a one-time legal charge that cut into profit.

What Meta reported for the third quarter

According to Meta's investor-relations press release, third-quarter 2026 revenue reached $64.601 billion, a 26% increase year over year, beating the roughly $63.3-63.4 billion Wall Street had expected, Fortune and Yahoo Finance reported. Net income came to $9.86 billion, or $3.83 per diluted share, up from about $2.7 billion and $1.05 per share a year earlier, Fortune reported.

That profit growth happened alongside a large legal charge. Meta's own press release reported a $9.89 billion charge tied to an August 2026 agreement with attorneys general in 52 U.S. states over teen mental-health litigation; Fortune reported the charge at approximately $10 billion, with about $18 billion to be paid over the next decade. Meta separately faces a $25 billion penalty ordered by a New Mexico judge over Cambridge Analytica-related data-privacy claims, which Meta plans to appeal, according to Fortune. Partly as a result, free cash flow turned negative at about -$2.1 billion, down from positive cash flow the prior quarter, Fortune reported. Meta's headcount stood at 66,775, down 15% year over year, per the company's own results.

Raising the AI spending floor

Meta's press release narrowed its full-year 2026 capital-expenditure guidance to a range of $135-145 billion, up from a prior floor of $130 billion, with total 2026 expenses now projected at $175-179 billion. Third-quarter capital expenditures alone, including finance-lease payments, came to $34.9 billion. Meta said the spending funds chips and data centers for training large AI models as well as development of its consumer AI agent, Muse, Yahoo Finance and the company's own materials indicated.

Meta also said it would create a new non-voting Class C share class, a structure management described as a way to keep raising capital for AI infrastructure while preserving founder Mark Zuckerberg's voting control, according to Yahoo Finance. In a statement accompanying the results, Zuckerberg pointed to progress on the company's AI products during the quarter, according to materials carried by multiple outlets.

Yahoo Finance also reported that Zuckerberg said Meta expects Muse's compute-cost efficiency to improve roughly tenfold within the next year, and that continued efficiency gains would make Muse profitable over multiple years. Meta's stock rose about 3% in after-hours trading following the release, Yahoo Finance reported.

A sector-wide pattern, not just Meta

Meta is not alone in raising its AI capital-expenditure guidance this year. Microsoft's fiscal third-quarter 2026 capital expenditures reached $30.9 billion, an 85% increase from $16.7 billion a year earlier, according to an analysis published by Global Data Center Hub. The same analysis reported that Microsoft disclosed a calendar-year 2026 capex plan of about $190 billion, exceeding analyst consensus of roughly $154.6 billion, while Azure cloud revenue grew 40% in constant currency and Microsoft's AI business reached a $37 billion annual run rate, up 123% year over year.

A separate analysis published by Tech Insider reported that Alphabet raised its 2026 capital-expenditure guidance to $180-190 billion, up from a prior $175-185 billion range, directing the money toward Google Cloud and custom TPU chips. The same analysis estimated the combined 2026 capital expenditure of Microsoft, Meta, Alphabet and Amazon at roughly $725 billion, describing this total as an aggregate estimate rather than a single confirmed company disclosure.

Why guidance keeps rising: component costs, not just new buildings

Both the Global Data Center Hub and Tech Insider analyses attributed a significant share of the 2026 increases to rising component prices rather than simply adding more data-center capacity. Global Data Center Hub reported that Microsoft attributed about $25 billion of its capex increase to "component price inflation rather than additional capacity," citing pressure on GPU and memory pricing, and noted Microsoft's gross margin slipped to 67.6% from 68.7%, which it linked to accelerated depreciation and component inflation. Tech Insider's analysis similarly attributed higher component pricing - particularly DRAM and memory costs - as a driver of higher 2026 spending plans across Microsoft, Meta and Alphabet, though the only executive statement the analysis quoted directly on this point came from Microsoft's chief financial officer.

What the earnings reports show about the AI investment debate

The figures reported by Meta, Microsoft and Alphabet show a specific, numbers-based version of a question that has shadowed the technology sector for several quarters: whether the scale of AI infrastructure spending is matched by revenue growth. Meta's own results show both sides of that question in the same report - record quarterly revenue and a stock-price gain after the release, reported by Yahoo Finance, alongside negative free cash flow and a sharply higher spending floor extending into 2027, according to the company's press release. None of the companies' statements cited in their earnings materials constitute independent verification that the spending will generate matching returns; those are company projections and executive statements, not audited outcomes.