Big Tech Triples Down on AI Infrastructure Spending Despite Bubble Concerns

Meta, Google, and Microsoft Triple Down on AI Spending

Three of the biggest US tech giants -- Microsoft, Meta, and Google -- sent investors a blunt message when they reported quarterly earnings on Wednesday: Their lavish spending on AI infrastructure is only just getting started. Meta said that its capital expenditure would total between $70 billion and $72 billion this year, up from its previous lower forecast of $66 billion to $72 billion. Next year, Meta's chief financial officer Susan Li said that she expected the company's spending would be "notably larger." The social media giant's soaring investment matches its soaring revenue: Meta reported raking in $51.24 billion last quarter, up 26 percent year-over-year. CEO Mark Zuckerberg said the company would keep pouring money into infrastructure to meet rising demand for AI and to prepare for potential major breakthroughs in the technology. "There's a range of timelines for when people think that we're going to get superintelligence," Zuckerberg said on a conference call with analysts. "I think that it's the right strategy to aggressively front-load building capacity, so that way we're prepared for the most optimistic cases." Meta has moved aggressively to recruit AI talent in recent months, offering some researchers compensation packages worth hundreds of millions of dollars. The company also cut some 600 jobs last week in what it said was an effort to make its AI teams more efficient. Meta has reorganized its AI teams numerous times over the past eight months. Meta assured investors that its AI investments were already reaping rewards for the company, but didn't share many specifics. Meta did say AI was benefiting the company's ad business and virtual reality product lines, and predicted it would propel those divisions to new heights in the future. Google's parent company, Alphabet, said it expected its 2025 capital expenditures to be between $91 billion to $93 billion. Earlier this year, Alphabet estimated that number would be just $75 billion. Like at Meta, the increase in spending was matched with an increase in revenue. The tech giant said it earned a record $102.3 billion in the third quarter, up 33 percent from a year ago. Most of Alphabet's spending will likely be funneled into data centers and other artificial intelligence initiatives. Google said it earned $15.15 billion from its cloud business in the third quarter, a 35 percent increase from the same period in 2024. Gemini, Google's general purpose AI app, now has 650 million monthly active users, up from 450 million last quarter. (For comparison, OpenAI CEO Sam Altman recently said that ChatGPT has 800 million weekly users.) Microsoft reported revenues of $77 billion for the quarter ending on September 30, up 18 percent from a year ago. Its cloud business revenue was up 26 percent year-over-year. Its capital expenditures were $34.9 billion this quarter, with much of the investment going toward AI infrastructure. That figure is nearly $5 billion more than previously forecasted, and a 74 percent jump up from the same quarter a year ago.

AI Is the Bubble to Burst Them All

I talked to the scholars who literally wrote the book on tech bubbles -- and applied their test. AI may not simply be "a bubble," or even an enormous bubble. It may be the ultimate bubble. What you might cook up in a lab if your aim was to engineer the Platonic ideal of a tech bubble. One bubble to burst them all. I'll explain. Since ChatGPT's viral success in late 2022, which drove every company within spitting distance of Silicon Valley (and plenty beyond) to pivot to AI, the sense that a bubble is inflating has loomed large. There were headlines about it as early as May 2023. This fall, it became something like the prevailing wisdom. Financial analysts, independent research firms, tech skeptics, and even AI executives themselves agree: We're dealing with some kind of AI bubble. But as the bubble talk ratcheted up, I noticed few were analyzing precisely how AI is a bubble, what that really means, and what the implications are. After all, it's not enough to say that speculation is rampant, which is clear enough, or even that there's now 17 times as much investment in AI as there was in internet companies before the dotcom bust. Yes, we have unprecedented levels of market concentration; yes, on paper, Nvidia has been, at times, valued at almost as much as Canada's entire economy. But it could, theoretically, still be the case that the world decides AI is worth all that investment. What I wanted was a reliable, battle-tested means of evaluating and understanding the AI mania. This meant turning to the scholars who literally wrote the book on tech bubbles. In 2019, economists Brent Goldfarb and David A. Kirsch of the University of Maryland published Bubbles and Crashes: The Boom and Bust of Technological Innovation. By examining some 58 historical examples, from electric lighting to aviation to the dotcom boom, Goldfarb and Kirsch develop a framework for determining whether a particular innovation led to a bubble. Plenty of technologies that went on to become major businesses, like lasers, freon, and FM radio, did not create bubbles. Others, like airplanes, transistors, and broadcast radio, very much did. Where many economists view markets as the product of sound decisions made by purely rational actors -- to the extent that some posit that bubbles don't exist at all -- Goldfarb and Kirsch contend that the story of what an innovation can do, how useful it will be, and how much money it stands to make creates the conditions for a market bubble. "Our work puts the role of narrative at center stage," they write. "We cannot understand real economic outcomes without also understanding when the stories that influence decisions emerge." Goldfarb and Kirsch's framework for evaluating tech bubbles considers four principal factors: the presence of uncertainty, pure plays, novice investors, and narratives around commercial innovations. The authors identify and evaluate the factors involved, and rank their historical examples on a scale of 0 to 8 -- 8 being the most likely to predict a bubble.

AI is making Google and Meta even stronger and richer

So they're increasing spending on infrastructure to keep it that way When generative AI exploded into public view in late 2022, plenty of pundits predicted it would be bad news for the likes of Google and Meta as nimble AI-powered rivals found new ways to capture netizens' attention and monetize it. Those pundits appear to have been wrong, as the two tech giants on Wednesday used quarterly earnings announcements to reveal their revenue has surged thanks to AI. Sundar Pichai, CEO of Google's parent company Alphabet, told investors that the web giant's recently-released search update that presents results as prose written by AI instead of just a list of links "is already driving incremental total query growth" and "driving an expansionary moment for Search." It's also driving more revenue. Alphabet's advertising haul rose 12 percent year-over-year to hit $74.2 billion for Q3. Google Cloud is also benefiting from AI. Pichai pointed to faster customer growth as orgs sign for Google's AI smarts, bigger deals, and deeper engagements. Google Cloud revenue for Q3 was $15.15 billion, a 33 percent year-over-year jump. Alphabet revenue for the quarter was $102.34 billion - the company's first $100 billion quarter - representing 16 percent year-on-year growth. AI also helped Meta to a strong quarter. CEO Mark Zuckerberg pointed [PDF] to AI-powered recommendations leading to users spending more time on the company's social networks. "Our ads business continues to perform very well, largely due to improvements in our AI ranking systems," he added, before revealing "the annual run-rate going through our completely end-to-end AI-powered ad tools has passed $60 billion." Meta is also growing, fast, with Q3 revenue of $51.25 billion representing 26 percent year-over-year improvement. Both companies plan to spend a lot more on AI infrastructure. Meta previously forecast 2025 capital expenditure of $66 billion to $72 billion. It raised the floor to $70 billion, then predicted "notably larger" growth in spending during 2026. Zuckerberg said much of that spending will go on infrastructure needed to develop "superintelligence" - Zuck-speak for a powerful personal assistant he hopes to develop and deploy. The CEO told investors Meta needs to "aggressively frontload building capacity" for superintelligence, and that doing so won't risk overinvestment in infrastructure that goes unused. "We'll use the extra compute to accelerate our core business, which continues to be able to profitably use much more compute than we've been able to throw at it," he said. "And we're seeing very high demand for additional compute, both internally and externally. And in the worst case, we were just slow building new infrastructure for some period while we grow into what we build. The upside is extremely high for both our existing apps and new products and businesses that are becoming possible to build." Our core business continues to be able to profitably use much more compute Alphabet CFO Anat Ashkenazi said the company expects FY 2025 capital expenditure to be in the range of $91 billion to $93 billion, up from previous estimates of $85 billion. She said 60 percent of Q3's $24 billion capex went on servers, with the remainder spent on data centers and networking equipment. Investors liked what they heard from Alphabet, as its shares spiked six percent in after-hours trading. Meta shares sank seven percent, probably because its net income slipped 83 percent year-over-year due to coughing up a one-time income tax charge of $15.93 billion. ®

In the AI boom, not all capex is created equal

Another tech earnings season, another dizzying escalation of the artificial intelligence capex spending boom. The headlong growth in Big Tech's data centre spending plans has been notable all year, but this week brought an extra twist. Meta and Microsoft both now predict their 2026 spending increases will be even bigger than 2025, while Alphabet jacked up its capex again for the rest of this year. Following the giant chip and cloud computing deals that OpenAI has forged in recent weeks, it's easy to see all this as an undifferentiated and undisciplined dash for growth. Look a little deeper, though, and there are big differences in the level of risk and likely returns. Once Wall Street's AI fervour eases, investors will pay more attention in discerning which companies have the most sustainable capital spending. One key difference is near-term demand. Microsoft, for instance, had expected that demand for its AI capacity would exceed supply until at least the end of this year. This week, its executives predicted the supply shortage would run until at least the middle of next year. Microsoft also claims the duration of business it has already booked closely echoes the expected useful life of its AI chips, helping to match its AI costs and revenue. Meta, by contrast, is building more on spec. Pressed repeatedly by analysts this week on the purpose behind his company's booming capex, chief executive Mark Zuckerberg was studiously vague. Building new services that reach billions of people, he said, "is a huge muscle that Meta has developed" -- though he didn't explain what those future AI services might be. Investors are still willing to give Zuckerberg the benefit of the doubt. Meta's stock price is more than six times higher than it was at its nadir three years ago, when Wall Street lost confidence in his earlier bet on the Metaverse. But the 10 per cent dent to the shares on Thursday suggests patience is starting to wear. Another important differentiator is the nature of the AI demand that some companies are starting to report. The backlog of contracted business -- known as remaining performance obligations, or RPO -- is a critical indicator. But not all future revenue looks the same. When Oracle revealed last month that its RPO had jumped to $455bn, the news sent its shares soaring. It only sank in later that $300bn of that is linked to a five-year deal with OpenAI, which is still a long way from generating the demand for all that capacity or the cash flow to pay for it. This week, Microsoft said its RPO rose by half to nearly $400bn, even before counting a $250bn deal it announced with OpenAI. Its executives were at pains to point out that these bookings reflect demand from a wide variety of customers, with a weighted average duration of only two years. In other words, these contracts should generate revenue in short order. This points to two different AI capex booms: A relatively short-term one, tied to some discernible level of actual demand, and a longer-term, speculative one, tied to an almost religious belief in exponential growth. A third key factor is profitability. With data centres now filling up rapidly with AI servers and other relatively shortlived assets, depreciation charges are starting to surge. At Microsoft, depreciation and amortisation jumped to 16.8 per cent of revenue in the latest quarter, from 11.3 per cent a year before. Expense discipline is becoming critical, even amid the boom -- as many workers at Amazon found to their cost this week. That should bring closer attention to which companies are building the most cost-effective AI infrastructure, and which have sustainable pricing power. Google and Microsoft both registered stronger profit margins than Wall Street had expected in the latest quarter, while Meta's operating margin dropped by three percentage points, to 40 per cent. The defensibility of those margins will be one of the big stories of 2026. Google is betting heavily on its full "stack" of in-house AI technology, from AI chips, known as TPUs, to leading foundation models. Microsoft, meanwhile, has been overtly steering away from what could become the most competitive, low-margin part of the AI market -- renting out GPU capacity. Its loosening ties with OpenAI have raised some anxiety among investors. CEO Satya Nadella intimated, though, there will be plenty of more profitable opportunities. When Wall Street's AI party finally ends, decisions like these will help determine which companies dominate over the long haul.

Wall Street hates Meta's AI spending guidance raise. We don't

Meta Platforms shares were taking a beating in extended hours trading on Wednesday after management raised its expense guidance and took a massive tax charge. Revenue in the three months ended Sept. 30 climbed 26% year over year to $51.24 billion, easily outpacing the consensus estimate of $48.14 billion, according to LSEG. Adjusted earnings per share came in at $7.25 versus the $6.69 consensus, LSEG data showed. That earnings number does not include a nearly $16 billion, or $6.20 per share, one-time income tax charge due to the implementation of President Donald Trump's One Big Beautiful Bill Act. Bottom line Given the non-recurring nature of the charge and reassurances from CFO Susan Li on the post-earnings conference call, we don't think the stock's 7.5% decline is due to the tax change. Li said, "We continue to expect we will recognize significant cash tax savings for the remainder of the current year and future years under the new law, and this quarter's charge reflects the total expected impact from the transition to the new U.S. tax law." Rather, the pressure on the stock was almost certainly due to management's expense guidance raise for the remainder of 2025, with Li reiterating prior commentary that capital expenditure "dollar growth will be notably larger in 2026 than 2025, with growth primarily driven by infrastructure costs, including incremental cloud expenses and depreciation." Total expense growth is also expected to "grow at a significantly faster percentage rate in 2026 than 2025," management also noted on the release. META YTD mountain Meta Platforms YTD We understand the capex concerns as the market tries to figure out what the long-term return on these monstrous artificial intelligence-driven investments is. But management has a good handle on things, with optionality and the ability to adapt depending on how things play out. To this point, CEO Mark Zuckerberg said on the call, "The right strategy [is] to aggressively frontload building capacity. So, that way we're prepared for the most optimistic cases. That way, if superintelligence arrives sooner, we will be ideally positioned for a generational paradigm shift in many large opportunities." In AI terms, superintelligence is when computers become smarter than humans. Zuckerberg added, "If it [superintelligence] takes longer to achieve, then we'll use the extra compute to accelerate our core business, which continues to be able to profitably use much more compute than we've been able to throw at it. And, we're seeing very high demand for additional compute, both internally and externally. And, in the worst case, we would just slow building new infrastructure for some period while we grow into what we build." That Meta can profitably leverage even more compute than it already has should ease the minds of some investors when it comes to the massive spending. Meta is clearly going to find use for all the infrastructure it's building, one way or another. It's better to have it and not need it immediately than need it immediately and not have it. While there is always the potential for return on investment (ROI) to materialize more slowly than expected or at a lower rate, it should nonetheless be positive over time. Why we own it Meta Platforms dominates the world of targeted ads with excellent technology, and its strong user engagement makes it a great place to advertise. The company's scale provides the financial power and employee talent needed to pursue new growth avenues such as artificial intelligence. Competitors : Alphabet , TikTok, and Snap Weight in portfolio : 4.69% Most recent buy : Sept. 6, 2022 Initiated : May 29, 2014 With these internal safety nets in place, the team is right to front-load the spending as the potential opportunity is simply too big to miss, and the stakes are simply too high. When technology as consequential as AI comes around, it's either get with the program and fight to lead the way, or risk being disrupted by those who are willing to spend, which, at this point, is just about everyone. On the call, Zuckerberg further explained the internal benefits of building out more AI infrastructure. "The upside is extremely high for both our existing apps and new products and businesses that are becoming possible to build across Facebook, Instagram, and Threads. Our AI recommendation systems are delivering higher quality and more relevant content, which led to 5% more time spent on Facebook in Q3 and 10% on Threads. Video is a particular bright spot, with video time spent on Instagram up more than 30% since last year." Outside of the one-time earnings hit - which we think the Street is looking past - and the spending dynamics, there really wasn't much to take issue with in the third quarter report. In fact, everything else was fantastic. Revenue in both segments outpaced expectations, while the Family of Apps operating income came in better than expected, and the operating loss at Reality Labs was nearly $700 million less than expected. Sales were better than expected across all geographies. Engagement, as represented by Family Daily Active People, was well ahead of expectations, as was Family Average Revenue per Person. Free cash flow came up a hair short on the back of elevated capex, but was pretty much in line thanks to a solid beat on operating cash flow versus expectations. We're reiterating our Meta price target of $825 per share and discussing whether to upgrade the stock to our buy-equivalent 1 rating. We currently have it as a 2 rating, which means look to buy on a pullback. We have to consider that while down after the release, the stock was up 28% year to date as of Wednesday's close. Quarterly highlights Instagram reached 3 billion monthly active users, while Threads recently passed 150 million daily actives with Zuckerberg saying it "remains on track to become the leader in its category." Reels reached an annual revenue run rate of over $50 billion. On the call, Zuckerberg noted the "annual run rate going through [Meta's] completely end-to-end AI-powered ad tools has passed $60 billion." Meta AI has over 1 billion monthly active users, with the team seeing increased usage as the underlying models improve. New Meta Ray-Ban and Oakley smart glasses are selling well. Zuckerberg said the new Meta Ray-Ban display glasses "sold out in almost every store within 48 hours, with demo slots fully booked through the end of next month." Total ad impressions across all services increased 14% year over year, with Li saying it was "healthy across all regions, driven by engagement and user growth, particularly on video services." Average price per ad increased 10% year over year. Regarding cash returns to shareholders, Meta returned $3.2 billion to shareholders via share repurchases and another $1.3 billion via dividends. Guidance Meta expects current fourth-quarter revenue in the range of $56 billion to $59 billion, which even on the low end, easily surpasses the consensus expectation of $54.95 billion, according to LSEG. Management, as mentioned earlier, raised the lower end of its full-year capital expenditures forecast. They are now targeting between $70 billion and $72 billion, up from the prior range of $66 billion to $72 billion. This new range is above the $68.36 billion consensus estimates, according to FactSet. Meta also raised the lower end of its 2025 total expenses guidance to between $116 billion to $118 billion, up from the prior range of $114 billion to $118 billion. This, too, appears to be above expectations of $114.9 billion, according to FactSet. (Jim Cramer's Charitable Trust is long META. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.

Tech giants are spending big on AI in a bid to dominate the boom

Its spending growth in 2026 is poised to be "notably larger" than this year, the company said. Meta is seeking to compete with companies like OpenAI. On a call with analysts, Meta boss Mark Zuckerberg defended the firm's investments, saying he saw big opportunities ahead driven by AI, both in terms of new products and for honing its current business selling ads and feeding people content. "The right thing to do is accelerate this," he said, adding later: "We are sort of perennially operating the family of apps and ads business in a compute-starved state at this point." Google and YouTube owner Alphabet similarly raised its forecast for this year to $91bn to $93bn, up from an earlier outlook of $85bn in the summer, in the latest sign of its increasingly lofty spending goals, That estimate is nearly double the capital expenditures that the company reported for 2024. Microsoft's capital expenditures in the quarter through to 30 September, including on data centres, totalled $34.9bn, the company reported on Wednesday - a larger spending figure than analysts had expected, and up from $24 billion in the previous quarter. "We continue to increase our investments in AI across both capital and talent to meet the massive opportunity ahead," Satya Nadella, Microsoft's chief executive, said. Azure, the firm's cloud computing unit, and Microsoft's other AI products have a "real-world impact", Mr Nadella said. Exuberance among investors about massive AI spending has helped all three tech firms outperform the broader S&P 500 index. But Wall Street is also focused on whether these firms' investments are starting to yield tangible returns. The two things holding up the US economy in the last several months have been consumers and AI-related business investments, said Aditya Bhave, senior US economist at Bank of America. "To the extent that the latter remains strong, it's a bullish signal for GDP growth," he said.

Investors continue to sound the alarm on the inevitable burst of the AI bubble

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. Bottom line: Despite Big Tech pouring trillions into AI initiatives and building massive new data centers, the expected returns may never materialize. Analysts warn that the hype far outpaces reality, creating a precarious financial bubble that could have ripple effects across the broader economy. Lauren Taylor Wolfe is exploring new investment opportunities in a bullish market, but told CNBC she plans to steer clear of anything tied to artificial intelligence. As co-founder of activist investment firm Impactive Capital, Wolfe recently voiced concern over Wall Street's near-total fixation on AI among major technology corporations. Like many analysts and executives before her, Wolfe said we are "absolutely" living in an AI bubble that will eventually burst. She offered no timeframe or prediction of its magnitude, only warning that many stakeholders stand to lose money, echoing OpenAI CEO Sam Altman's prediction earlier this year. "Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes," Altman said in August. "That's not rational behavior. Someone's gonna get burned ... Someone is going to lose a phenomenal amount of money." Investment firms, Big Tech, and Wall Street continue betting big on the promise of unparalleled productivity gains from AI. Wolfe said an unprecedented amount of capital is currently earmarked for future AI initiatives, while the "Mag 7" are generating massive cash flow from their products and services - but still showing little return on investment. The Magnificent Seven refers to the market's most influential tech-driven companies - Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, and Tesla - which together make up a dominant portion of the S&P 500's total value. "There are trillions of dollars that are being earmarked to be spent relative to hundreds of billions of dollars of free cash flow generated by the Mag 7," she said. "They're going to have to borrow to invest in all this CapEx, and we have yet to see the returns on investment." Even worse, Big Tech's AI initiatives are unlikely to generate trillions in profits over the next five years. Simply put, the math does not support that kind of exponential return. Analysts have also warned that when the AI bubble finally bursts, the broader US economy could suffer as a result. That is why Wolfe and Impactive Capital are focusing on alternative investments to shield clients from another dot-com - style collapse. The firm is currently eyeing Advanced Drainage Systems, an "AI-proof" company specializing in stormwater and residential septic solutions.

AI investment the only thing keeping the US out of recession

Datacenter infrastructure and model development spending offset high borrowing costs AI spending is keeping the US economy out of recession, with datacenter infrastructure and model development providing the only significant growth amid trade turmoil, tariff shocks, and high borrowing costs. This is offsetting the drag from higher interest rates and the impact of the Trump administration's chaotic trade policies, numerous sources say. "AI has kept the economy out of a recession," BNP Paribas chief US economist James Egelhof told Yahoo Finance this week, noting the spending boom has convinced businesses that robust growth and a productivity surge are imminent. Apollo Global Management chief economist Torsten Sløk wrote yesterday: "There is basically no growth in corporate capex outside of AI at the moment." Unlike typical investment patterns, AI spending hasn't fallen despite Federal Reserve rate hikes because, Sløk noted, datacenter investment is ultimately financed by rising equity valuations of the "Magnificent Seven" including Microsoft, Amazon, Alphabet, and Nvidia. Analyst firm Omdia estimates that global datacenter capex will top $657 billion in 2025, almost double the figure of just two years ago, with the US dominating. Amazon's annual datacenter spending alone exceeds $100 billion, roughly Costa Rica's entire GDP. Jason Furman, Economist and former Deputy Director of the US National Economic Council, estimated in a New York Times podcast that 92 percent of the economic demand in the first two quarters of this year came from information processing equipment and software. "Ultimately, what we're really hoping for is that AI shows up on the supply side of the economy, actually helping us do more with less," he said. Some of that is happening, he claimed, "but there hasn't been anything particularly special about productivity growth to date." This is echoed in numerous studies showing that despite the billions being poured into AI, the return on investment remains uncertain. A UK government trial of Microsoft's M365 Copilot found no discernible gain in productivity, speeding up some tasks yet making others slower. US research revealed that companies invested $35-40 billion in generative AI initiatives, yet 95 percent have seen zero returns. Questions about unsustainability are intensifying. Management consultants at Bain & Company estimate that current spending trajectories would require the tech sector to generate $2 trillion in annual AI sales by 2030 - fueling concerns about another tech bubble. The Bank of England's Financial Policy Committee warned this month about the growing danger of a sudden correction in the financial markets due to inflated tech and AI stock valuations. Financial analysts have expressed particular concern over OpenAI's $500 billion valuation despite the AI darling not turning a profit yet. OpenAI chief Sam Altman himself acknowledged the AI industry is in a bubble, but seemed relaxed about potential consequences, even though a recent report claimed the company is losing about three times more money than it's earning - with just 5 percent of ChatGPT's 800 million users actually paying for it. ®

Zuck Acknowledges Meta Might Not Use AI Investments for New AI After All

You know how Meta, the company formerly known as Facebook, is spending $72 billion on AI infrastructure in 2025? Well, with his company’s stock in the middle of a vertigo-inducing plunge of about 8% on Wednesday afternoonâ€"a signal that investors are getting worriedâ€"CEO Mark Zuckerberg conjured a vision of a world where Meta doesn’t end up using all that stuff for its intended purpose in the near future, saying it would be fine. That infrastructure Meta is investing in is intended to come in handy if AI superintelligence is achieved soon, Zuckerberg said on an earnings call Wednesday. “If it takes longer, then we’ll use the extra compute to accelerate our core business, which continues to be able to profitably use much of the compute,†Zuckerberg said, according to The Wall Street Journal. Meta’s “core business†remains ad-subsidized social media apps. Advertising accounts for almost all of its revenue, according to its own financial disclosures from earlier this year. It’s not as though Meta has not deployed AI in its social media apps. (Who can forget the tragic case of “Big sis Billieâ€?) But the basic functions of Meta’s platforms have not exactly been revolutionized by AI, unless you count all those people who now post things like Shrimp Jesus on Facebook. With that in mind, might billions of dollars spent on data center construction, expensive AI experts, and an entire GPU company be overkill, assuming it's just to "accelerate" Facebook and Instagram? And those AI expenses aren’t about to shrink next year. Zuckerberg also told investors to expect 2026 to be even spendier than 2025. “I think it’s the right strategy to aggressively front-load building capacity,†he said, according to the Journal “.That way, if superintelligence arrives sooner, we will be ideally positioned for a generational paradigm shift in many large opportunities.†The worst case scenario for Meta, Mark Zuckerberg says, is that it might have to “slow building new infrastructure for some period while we grow into what we build.†Gizmodo reached out to Meta for comment, and will update if we hear back.

Big Tech tests investors' patience with $80bn AI investment spree

Google, Meta and Microsoft spent almost $80bn over the last quarter on artificial intelligence infrastructure, but investors had markedly different reactions to their plans to increase this historic spending spree. Alphabet shares rose almost 7 per cent in after-hours trading on Wednesday as the Google parent boosted its capital expenditure plans for 2025 by $8bn to $93bn, and delivered a record $100bn in quarterly revenue. By contrast, Meta plunged almost 9 per cent -- potentially wiping $160bn from its valuation when markets open on Thursday -- as Mark Zuckerberg's company signalled much higher AI spending, which could top $100bn next year. The varied reaction to their earnings and spending plans "underscores how sensitive investors are to how quickly the AI build-out can deliver revenue", said Dec Mullarkey, managing director of $300bn asset manager SLC Management. "Investors are worried that the rush to grab market leadership may cause an overshoot," he added. "No one needs reminding that history is full of episodes of technology exuberance that eventually left the early investors battered." Microsoft -- which became the third company to surpass a $4tn valuation this week after finalising its restructuring pact with OpenAI -- also suffered a share price drop. Its stock fell 4 per cent, despite beating profit estimates and posting a 39 per cent jump in revenue at its key Azure cloud computing unit. It reported capital expenditure was $35bn in the quarter, a 74 per cent increase year-on-year and $5bn more than expected. Executives forecast spending of almost $140bn next year. Chief executive Satya Nadella told analysts that the software group was building "planet scale" cloud infrastructure and plans to double Microsoft's data centre footprint over the next two years. Google and Microsoft, which both sell cloud computing to other businesses, had an easier time showing investors that elevated spending on chips, data centres and electricity will lead to income. After a slow start in the AI race, chief executive Sundar Pichai said that the Gemini App, its main consumer AI product, now has 650mn monthly users, up from 450mn in July, and closing in on ChatGPT's 800mn. Pichai added that growth in its cloud unit "was driven by enterprise AI products, which are generating billions in quarterly revenues" and that it had an order backlog for computing services worth $155bn. The 15 per cent boost to core search advertising revenue also helped address concerns that ChatGPT is taking market share and AI is cannibalising traditional search. "We believe this performance demonstrates successful AI integration across ad-based platforms," said Angelo Zino, an analyst at CFRA Research. "Google's ability to maintain margins while scaling AI infrastructure demonstrates effective use of spending." Zuckerberg, meanwhile, had to defended huge spending on infrastructure for Meta's own use, as the tech group vies to be the first to build artificial superintelligence. He said it was "the right strategy to aggressively frontload building capacity". He added that any excess data centre space could be repurposed to serve Meta's core advertising functions, which he said were "compute starved". A 26 per cent increase in quarterly revenue to $51.2bn failed to mollify the market, as investors fretted that Meta's huge outlay on chips and staff has yet to produce a large language model as capable as rivals. The social media company said capex could hit $72bn by year-end and that spending growth would be "notably larger" in 2026, implying a number far in excess of an earlier forecast for $105bn. Zuckerberg has also been luring engineers to his elite "TBD" lab with pay packages in the hundreds of millions of dollars, which Meta warned would be a big contributor to expenses as full-year costs appear in its results. Investors were disappointed by a rise in research and development costs, which accounted for 30 per cent of revenue, the highest level in more than two years. Its operating margin narrowed 3 percentage points to 40 per cent. "Expenses are growing faster than revenue," said Gene Munster at Deepwater Asset Management. "Next year it's going to be more like 18 per cent revenue growth and 35 per cent expense growth." Meta disclosed a $15bn one-off charge related to changes in President Donald Trump's tax bill that depressed net income 83 per cent to $2.7bn. Meta has indicated that its AI efforts are unlikely to generate meaningful revenue this year or in 2026. Zuckerberg on Wednesday promised that his new superintelligence team were focused on "novel" work that could be rapidly rolled out to 3.5bn users on Facebook, WhatsApp and Instagram, and could make money via advertising, commerce or subscriptions. Investors worry Zuckerberg's quest to dominate advanced AI is disconnected for Meta's underlying business despite his insisting that it can improve advertising ranking and recommendations. Brian Wieser, an analyst at advisory firm Madison and Wall, said Google and Microsoft "are doing much more from a tech perspective. Meta's actual business is selling ads." "There [are] so many more arrows in the quiver for Google and Microsoft," he added.

Tech giants pledge to spend more on AI despite bubble concerns

Meta CEO Mark Zuckerberg said on Wednesday that his company would increase its spending on artificial intelligence, despite the risk of over-investment. (Jeff Chiu/AP) SAN FRANCISCO -- Google and other large tech firms have taken an outsize role in the U.S. economy this year, becoming major drivers of economic growth by spending hundreds of billions on building data centers to power artificial intelligence software. Now the sector's economic sway is poised to grow further, after Google, Meta and Microsoft said on Wednesday that they would spend more than previously expected on AI infrastructure. The announcements, made in quarterly earnings releases, came hours after the Federal Reserve cut interest rates, citing an uncertain job market. Tech giants' pledges to increase AI investment come after industry leaders including Meta CEO Mark Zuckerberg have said the sector has inflated a financial bubble around the technology that could burst, causing a downturn for tech firms that could ripple across the economy. The revised spending plans mean that Google, Meta, Microsoft and Amazon are set to spend nearly $370 billion this year on construction of data centers, electricity-guzzling facilities packed with powerful computer chips used to create and run AI software such as ChatGPT. Google chief financial officer Anat Ashkenazi said on an earnings call Wednesday that the company would spend between $91 billion and $93 billion this year on AI, an increase from its previous estimate of $85 billion. Meta said that its AI spending would be at the high end of its previous estimate of $66 billion to $72 billion and "notably larger" next year, in commentary released with its earnings report on Wednesday. Microsoft said in its own report that it also expects its AI spending to grow in 2026, increasing from its anticipated outlay of $80 billion this year. Recent investments into AI data centers and business spending on AI software dwarf previous tech booms such as the initial investment into internet infrastructure. By translating into more business for not just the tech sector, but also construction, trucking and energy firms, AI investment has delivered an economic shot in the arm that has helped boost U.S. economic growth despite flagging consumer spending and a softening labor market. But the spending spree has fueled concerns among economists and tech leaders that the hopes for huge profits from AI products have created a massive financial bubble. Despite the popularity of apps such as OpenAI's ChatGPT, the chatbot and many other recent AI ventures remain unprofitable. A pullback on AI spending and expectations for the technology could harm tech companies dependent on speculative investment, undermine businesses across the nation that have been buoyed by data center construction and wipe billions off the value of tech stocks sitting in retirement accounts. OpenAI chief executive Sam Altman and Meta's Zuckerberg have committed to huge spending on AI infrastructure, but both have also said they believe the industry has created an investment bubble that could lead to financial losses for some companies. The Washington Post has a content partnership with OpenAI. Zuckerberg said on Meta's earnings call on Wednesday that the company's drive to build "superintelligence," a term for machines hoped to one day outperform humans in every way, makes the risk of overinvestment worthwhile. "Some people think that we'll get there in a few years. Others think it will be five, seven years or longer," Zuckerberg said. "I think that it's the right strategy to aggressively front load building capacity, so that way we're prepared for the most optimistic cases. "In the worst case, we would just slow build new infrastructure for some period while we grow into what we build," Zuckerberg said. A major slowdown in data center construction would have knock on effects outside of the tech industry including in the energy sector, where the facilities' power demands have increased electricity consumption and brought new investment. On Tuesday, Google said it would reopen a defunct nuclear power plant in Iowa to provide more electricity to power its facilities. The deal follows one struck by Microsoft last year to reopen the Three Mile Island nuclear plant in Pennsylvania. This week, OpenAI asked the White House to double the amount of new electrical power generation the United States builds each year, in order to compete with China's ability to provide more energy to its own AI industry. Investors continue to bet that the tech industry's huge expectations for AI will be met. Nvidia, the computer chip company whose products are instrumental to training AI algorithms, became the first company to reach a market value of more than $5 trillion Wednesday. It's stock rose after President Donald Trump mused ahead of his meeting with Chinese president Xi Jinping that he may ease restrictions on sales of the company's most powerful chips in China. Wednesday's announcements suggest investors will continue to back companies driving and benefiting from the AI investment boom, Gene Munster, managing partner of Deepwater Asset Management, said in a video posted on X. "These companies are continuing to talk about spending and investing much more than we thought three, six months ago," he said. "The AI trade is intact."

Meta CEO Mark Zuckerberg defends AI spending: 'We're seeing the returns'

Mark Zuckerberg, chief executive officer of Meta Platforms Inc., during the Meta Connect event in Menlo Park, California, US, on Wednesday, Sept. 17, 2025. Meta CEO Mark Zuckerberg is sounding a familiar tune when it comes to artificial intelligence: better to invest too much than too little. On his company's third-quarter earnings call on Wednesday, Zuckerberg addressed Meta's hefty spending this year, most notably its $14.3 billion investment in Scale AI as part of a plan to overhaul the AI unit, now known as Superintelligence Labs. Some skeptics worry that the spending from Meta and its competitors in AI, namely OpenAI, is fueling a bubble. For Meta's newly formed group to have enough computing power to pursue cutting-edge AI models, the company has been building out massive data centers and signing cloud-computing deals with companies like Oracle, Google and CoreWeave. Zuckerberg said the company is seeing a "pattern" and that it looks like Meta will need even more power than what was originally estimated. Over time, he said, those growing AI investments will eventually pay off in a big way. "Being able to make a significantly larger investment here is very likely to be a profitable thing over, over some period," Zuckerberg said on the call. If Meta overspends on AI-related computing resources, Zuckerberg said, the company can repurpose the capacity and improve its core recommendation systems "in our family of apps and ads in a profitable way." Along with its rivals, Meta boosted its expectations for capital expenditures. Capex this year will now be between $70 billion and $72 billion, compared to prior guidance of $66 billion to $72 billion, the company said. Meanwhile, Alphabet on Wednesday increased its range for capital expenditures to $91 billion to $93 billion, up from a previous target of $75 billion to $85 billion. And on Microsoft's earnings call after the bell, the software company said it now expects capex growth to accelerate in 2026 after previously projecting slowing expansion. Alphabet was the only one of the three to see its stock pop, as the shares jumped 6% in extended trading. Meta shares fell about 8%, and Microsoft dipped more than 3%. Zuckerberg floated the idea that if Meta ends up with excess computing power, it could offer some to third parties. But he said that isn't yet an issue. "Obviously, if you got to a point where you overbuilt, you could have that as an option," Zuckerberg said. In the "very worst case," Zuckerberg said, Meta ends up with several years worth of excess data center capacity. That would result in a "loss and depreciation" of certain assets, but the company would "grow into that and use it over time," he said. As it stands today, Meta's advertising business continues to grow at a healthy pace thanks in part to its AI investments. "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 under investing," Zuckerberg said. Revenue in the third quarter rose 26% from a year earlier to $51.24 billion, topping analyst estimates of $49.41 billion and representing the company's fastest growth rate since the first quarter of 2024.

The AI boom isn't going anywhere

Driving the news: Meta, Microsoft and Google -- some of the major "hyperscalers" driving the AI transformation -- all made bullish comments Wednesday on their spending plans. * Meta raised its spending forecast, saying its capital expenditures on AI infrastructure and the like will be at least $70 billion this year, and "notably larger" next year. * Google parent Alphabet -- fresh off a record $100 billion revenue last quarter -- raised its own spending forecast for the year to at least $91 billion. * Microsoft CEO Satya Nadella said strong demand was the reason they "continue to increase our investments in AI across both capital and talent." Between the lines: Tech giants and their investors are thrilled by all the new business. * Still, some of the discussion in this week's earnings calls suggests that demand is coming from companies spending beyond their means or financing each other in a loop that could unravel if one link breaks. Zoom out: The AI boom is so large, and at this point so self-sustaining, it's taken on an economic life of its own. * Federal Reserve chair Jerome Powell, during a news conference Wednesday, rejected the idea that the Fed lowering the cost of money would fuel an AI bubble in some way. * "I don't think that the spending that happens to build data centers all over the country is especially interest sensitive," Powell said. "It's based on longer-run assessments that this is an area where there's going to be a lot of investment that's going to drive higher productivity and that sort of thing." * There's little doubt about the ongoing impact of all these hundreds of billions of dollars in spending. * "This has been an important backstop for the economy and without which we would have seen substantially weaker growth numbers," Vanguard global chief economist Joe Davis wrote Wednesday. Zoom in: The ongoing evidence is clear from companies like Caterpillar, as Axios' Nathan Bomey writes. * Caterpillar CEO Joe Creed said on an earnings call that sales of equipment in the company's power generation segment soared 33%, "primarily due to demand for reciprocating engines for data center applications." * AI is lifting boats beyond the chipmakers, and fueling insatiable demand all up and down the industrial supply chain. The intrigue: The boom is boosting bottom lines and driving stock market records, but not necessarily translating into jobs yet. * As Powell noted Wednesday, the labor market continues to soften. * Data centers are good for construction jobs in the short term, but generally don't require huge staffing once they're built. * Companies like AI heavyweight Nvidia say they need more talent and will keep growing, but it's not clear whether that will be enough to offset signs of rising corporate layoffs. The bottom line: The race to build the future of AI isn't anywhere near over, and Wednesday's earnings show that the finish line still isn't in sight.

Silicon Valley called -- the 1990s are back

San Francisco has been economically detached from the rest of America since the internet boom of the mid-1990s. As everything from housing rents to per capita growth rates far higher than the US national average show, the Bay Area exists in its own orbit. I couldn't help but think of this when I flew into San Francisco a couple of weeks ago. Artificial intelligence, rather than consumer web pages, drives the exuberance today. But the vibe is the same as it was 30 years ago -- there is one story, and everyone is buying it. Every single billboard I passed from the airport to the city had something to do with AI. Giant white blocks of brand-new housing and office space lined the highway. In the windows, I saw young men (Silicon Valley employees are still mostly men, just like back then) typing away at computers while sports or video games played on their oversized flatscreen televisions. So far, so 1990s. Which leads me to wonder, what's different this time around? It's a topic that has been much discussed in recent months as it has become clear how much of the US growth picture depends on AI. While some of the answers are easy -- most people agree that the long-term impact of AI will be far broader and deeper than the rise of the consumer internet, even if we see a short-term correction -- I would point to three comparisons between the 1990s and today that warrant more investor consideration. First, AI is highly capital intensive. The dotcom bubble wasn't. Back then, anyone could throw up a homepage with almost no investment. Today, AI investment represents all the growth in corporate America's capital expenditure. Contrarians might point out that there were more than 80mn miles of fibre optic cable laid from the mid-1990s until the end of the dotcom boom, and much of that investment took years to pay off. US telecom companies spent $444bn in capital expenditure between 1996 and 2001. Still, compare this with the $342bn that will be spent this year alone in the US by the top investors in AI data centres and computing infrastructure, including Microsoft, Alphabet, Amazon and Meta. At the current rate of power consumption needed to fuel AI development, estimated investment will stretch to nearly $7tn by 2030. This brings us to another key point on the capital investment front, which is that energy is a hugely significant constraint on the development of AI today. That was not true back in the 1990s. If the current energy economics of AI hold, then the investment story becomes even more costly and complex, since it is dependent on the trajectory of grid upgrades, more nuclear power coming online, how quickly renewables can replace fossil fuel and how high oil prices will go amid these changes. Another difference between the two eras is that AI investment today is increasingly funded not by debt, but by equity. At first glance, that seems like a very good thing. The half a trillion dollars in telecom investment during the dotcom bubble was funded mostly by debt, whereas the big AI players today are huge, profitable companies. For that reason, we probably won't see a repeat of failures like WorldCom or Global Crossing, since the Magnificent Seven (or perhaps four, when it comes to AI) are not about to go broke. The richest companies in the history of global capitalism have cash to burn. But this leads to what may be the most important comparison between the 1990s and today. Many market participants argue that we shouldn't worry so much about the fact that AI hype seems to be holding up the entire US equity market, since the earnings of the Big Tech companies driving the boom are so huge. Yet when you look beneath the hood, that's not the full story. Apollo's chief economist Torsten Slok, who is known for his ability to land on the metrics that really matter, put out a remarkable chart a few days ago. It showed that since Donald Trump's tariff "liberation day" in April, companies in the Russell 2000 with negative earnings are outperforming those with positive earnings. This small-cap index includes many companies in AI-investment heavy areas such as software, biotechnology and healthcare. But it also includes plenty of others in sectors such as financials, energy, materials and communications, which are similarly betting on an AI future. It's an uncomfortable rhyming of history with the dotcom boom, when negative-earnings companies also overtook those with positive earnings for a prolonged period. What should we take from all this? A few things. First, although "AI will be a revolution in so many ways, there is an echo effect from the 1990s that we should pay close attention to", says Slok. What is more, the economic impact of even a short-term collapse in the AI bubble may be bigger than it was in the dotcom era, given that more Americans now hold more stock than ever before. Second, the dynamics underlying this bubble are far more complex than they were two decades ago. Will productivity gains offset AI-related job losses? Will energy remain a chokepoint? Will a Chinese innovator disrupt the market? Or will US growth continue to outpace the world, as it did in the 1990s? Prospects for more than just Silicon Valley hinge on the answers.

Big Tech's next earnings test: power and patience

Earnings week for Big Tech now reads like a utility story wearing a software badge. Five of the companies that set the market's temperature will step up to the plate this week with the same through-line: They've spent heavily to stand up AI capacity, and now they have to show it's pulling its weight. Big Tech has spent the past year constructing the future in real time -- pouring billions into data centers, power lines, and silicon pipelines as if tomorrow were already on the clock. The AI boom was once framed as a software story, but the real action now lives underground and under voltage. As Microsoft, Alphabet, Meta, Amazon, and Apple prepare to report, the market's fascination with AI is becoming something more material: a question of who can turn infrastructure into income before the meter runs out. Analysts are still treating this as a victory lap. Wedbush called this week a "validation moment" for the sector, projecting a combined three-year wave of nearly $3 trillion in AI-driven enterprise and government spending. They describe it as a "1996 moment, not 1999" -- the start of a structural boom, not a speculative bubble. The Street's collective faith is that capital expenditure is no longer a cost but a competitive moat, and that the companies spending fastest are the ones least likely to fall behind. It's a compelling story -- right up until someone misses a delivery window. A warning may have come this week from Super Micro Computer, the unglamorous backbone of the AI supply chain. The California server-maker cut its quarterly forecast from as much as $7 billion in revenue to about $5 billion, blaming "delayed shipments." Demand was fine, executives insisted -- it was everything else that wasn't on time. The company reaffirmed its full-year guidance, but the revision landed like a speed bump in a market that has learned to treat logistical friction as heresy. Investors have been willing to underwrite billions in unfinished projects on the assumption that delivery is just a formality. Super Micro's stumble reminded them that time isn't always a variable you can buy when it comes to an economy running on electricity, concrete, and copper. The stakes this week are enormous because Microsoft, Apple, Amazon, Alphabet, and Meta play a massive role in shaping the market. Together, they account for roughly 45% of the Nasdaq 100's value and a quarter of the S&P 500's, meaning the direction of U.S. equities now depends on their ability to convert capex into earnings. Meta, once cautious, now expects to spend up to $72 billion next year on data-center retrofits and AI infrastructure. Amazon isn't far behind, pairing record AWS investment with a steady surge in its ad business, which has quietly become a $15-billion-per-quarter profit engine. Even Apple, traditionally allergic to infrastructure narratives, has nudged its own capital budget higher as it tiptoes into on-device AI. To hear analysts tell it, this is not reckless. It's foundational. The view from Wedbush and Deutsche Bank is that the current spending binge is still in its early innings, the monetary scaffolding for a generation-long shift toward AI-infused everything. But those same reports also concede that returns will lag, and that power and materials are now the limiting reagents in Big Tech's growth chemistry. Every time a company describes "temporary capacity constraints," it might just be corporate code for having built faster than the grid can carry. It's a strange equilibrium, this strength that is showing up as stress. Every hyperscaler insists the physics problem is temporary (transformer shortages and interconnection delays are pushing timelines and costs in key grids), yet none can outrun it. The infrastructure cycle that started as an arms race now looks more like a trench war -- measured in megawatts instead of market share. Power hookups are delayed, transformers are back-ordered, and the median data-center project takes months longer than projected to move from shell to service. The shortage of time is creating its own economics: server makers report backlog growth, utilities raise connection fees, and cloud providers find themselves paying peak-hour rates just to keep workloads online. Meta, perhaps, illustrates that paradox most clearly. The company's ad engine still throws off extraordinary cash flow, enough to fund its AI transition without borrowing, but the math is beginning to look less like a software business and more like a utility build-out. The company's entire bet rests on proving that all this power and infrastructure will translate into smarter targeting and richer margins in the quarters ahead. So far, investors have rewarded the promise, not the proof. By midweek, attention will shift to Meta, whose transformation from ad platform to infrastructure player still seems to baffle and fascinate the Street in equal measure. Every update reads like a personality test: optimists call it evolution, skeptics call it wandering, and everyone agrees the power bill keeps climbing. Amazon rounds out the cloud picture on Thursday, and the mood there has subtly improved. Last quarter's AWS slowdown lowered expectations, and analysts such as Wedbush's Scott Devitt now frame 2025 as a bridge year, with a "breakout" coming in 2026 as backlog converts to revenue. The offsetting comfort is advertising, where operating margins remain high enough to subsidize AI infrastructure without denting profits. Then Apple, closing out the week, offers the control sample -- a company still defined by consumer timing, not data-center timing. The iPhone 17 cycle is running about 14% ahead of its predecessor's, and services are expected to continue to deliver margins north of 40%. Apple's earnings will serve as the macro mirror: If its consumers keep spending, the rest of the tech ecosystem can keep pretending the power bill is someone else's problem. Because this is a preview, not a post-mortem, this week's tells will be language and sequencing. On the cloud names, listen for specificity around capacity timing -- not "we're building," but "these sites are now online; this many racks are productive; here's what that means for margin in Q4 and FY26 Q1." On ads, investors will be looking for clean explanations of what AI is doing to yield -- fewer poetic flourishes, more mechanics. On Apple, the tone of the holiday guide often matters more than a tenth of a point on gross margin; a confident Services arc can cover a lot of short-term noise. The larger thread here will be patience. These companies have collectively asked investors to accept a lag -- cash out the door now, capacity online later, revenue ramping after that -- in exchange for a fatter, stickier earnings base on the other side. This week is where that bet gets measured in real time across five different business models. Either way, this week compresses a year's worth of AI promises into 48 hours of earnings calls. By Friday morning, we'll know whether the power story is still a tailwind -- or whether the grid has started to set the pace for the Street. The week's most likely outcome is neither triumph nor reckoning, but another week of qualified validation: solid revenue, squeezed margins, and promises that the payoff is coming. The market will call that success, because the alternative is to admit how much of the boom still lives on construction sites. The faith underwriting this entire cycle isn't just in AI itself; it's in the belief that time is elastic -- that the delivery can always catch up to the build. For now, investors are still buying that story. The revolution continues to advance on schedule, at least on paper, even if the substations haven't caught up. But patience, like power, has a cost. And this week, the most valuable companies in history will have to prove they can keep paying both.

Google expects 'significant increase' in capital expenditure in 2026, execs say

Sundar Pichai, chief executive officer of Alphabet Inc., during the Bloomberg Tech conference in San Francisco, California, US, on Wednesday, June 4, 2025. Google parent Alphabet is planning a "significant increase" in spend next year as it continues to invest in AI infrastructure to meet the demand of its customer backlog, executives said Wednesday. The company reported its first $100 billion revenue quarter on Wednesday, beating Wall Street's expectations for Alphabet's third quarter. Executives then said that the company plans to grow its capital spend for this year. "With the growth across our business and demand from Cloud customers, we now expect 2025 capital expenditures to be in a range of $91 billion to $93 billion," the company said in its earnings report. It marks the second time the company increased its capital expenditure this year. In July, the company increased its expectation from $75 billion to $85 billion, most of which goes toward investments in projects like new data centers. There'll be even more spend in 2026, executives said Wednesday. "Looking out to 2026, we expect a significant increase in CapEx and will provide more detail on our fourth quarter earnings call," said Anat Ashkenazi, Alphabet's finance chief. The latest increases come as companies across the industry race to build more infrastructure to keep up with billions in customer demand for the compute necessary to power AI services. Also on Wednesday, Meta raised the low end of its guidance for 2025 capital expenditures by $4 billion, saying it expects that figure to come in between $70 billion and $72 billion. That figure was previously $66 billion to $72 billion. Google executives explained that they're racing to meet demand for cloud services, which saw a 46% quarter-over-quarter growth to the backlog in the third quarter. "We continue to drive strong growth in new businesses," CEO Sundar Pichai said. "Google Cloud accelerated, ending the quarter with $155 billion in backlog." The company reported 32% cloud revenue growth from the year prior and is keeping pace with its megacap competitors. Pichai and Ashkenazi said the company has received more $1 billion deals in the last nine months than it had in the past two years combined. In August, Google won a $10 billion cloud contract from Meta spanning six years. Anthropic last week announced a deal that gives the artificial intelligence company access to up to 1 million of Google's custom-designed Tensor Processing Units, or TPUs. The deal is worth tens of billions of dollars. The spend on infrastructure is also helping the company improve its own AI products, executives said on the call. Google's flagship AI app Gemini now has more than 650 million monthly active users. That's up from the 450 million active users Pichai reported the previous quarter. Search also improved thanks to AI advancements, executives said. Google's search business generated $56.56 billion in revenue -- up 15% from the prior year, tempering fears that the competitive AI landscape may be cannibalizing the company's core search and ads business. AI Mode, Google's AI product that lays within its search engine, has 75 million daily active users in the U.S., and those search queries doubled over the third quarter, executives said. They also reiterated that the company is testing ads in that AI Mode product.

The cost of AI, high; the cost of sucking up to Trump, higher. Meta's Marc Zuckerberg crunches some big numbers

The cost of AI expansion remains high and the price of sucking up to President Trump even higher as Meta revealed in its Q3 numbers. Despite CEO Marc Zuckerberg's oleaginous efforts to get back in the favors of the Trump 2.0 administration, Meta's profits took a pummelling from a $16 billion charge related to the so-called Big Beautiful Bill. That took profit for the quarter down from $18.64 billion to $2.71 billion. Meanwhile revenue growth is now outstripped by cost growth, the former running at 26% year-on-year, the latter at 32%. But Zuckerberg committed again to spending "notably larger" sums on AI rollout, with Meta raising CapEx estimates for this year to between $70 billion and $72 billion this year, compared with its prior forecast of $66 billion to $72 billion. It has to be done, he argues: We're building what we expect to be an industry-leading amount of compute. Now there's a range of time lines for when people think that we're going to get super-intelligence. Some people think that we'll get there in a few years. Others think it will be 5, 7 years or longer. I think that it's the right strategy to aggressively front-load building capacity so that way we're prepared for the most optimistic cases. That way, if super-intelligence arrives sooner, we will be ideally positioned for a generational paradigm shift in many large opportunities. And in words to chill the hearts and wallets of Wall Street short-termists, he left the company wiggle-room to allow for further rises in spending: We keep on seeing this pattern where we build some amount of infrastructure to what we think is an aggressive assumption. And then we keep on having more demand to be able to use more compute, especially in the core business in ways that we think would be quite profitable, then we end up having compute for. So I think that suggests that being able to make a significantly larger investment here is very likely to be a profitable thing over some period because if the primary use of it is going to be to accelerate the AI research and the new AI work that we're doing and how that relates to both the core business and new products. It won't be money down the drain, he added: Any compute that we don't need for that we feel pretty good that we're going to be able to absorb a very large amount of that to just convert into more intelligence and better recommendations in our family of apps and ads in a profitable way. Now I mean, it's, of course, possible to overshoot that, right? And if we do, then we see that there's just a lot of demand for other new things that we build internally, externally. Almost every week, people come to us from outside the company asking us to stand up an API service or asking if we have different compute that they could get from us and we haven't done that yet. But obviously, if you got to a point where you ever built, you could have that as an option. And then the kind of the very worst case would be that we effectively have just pre-built for a couple of years, in which case, of course, there would be some loss and depreciation, but we'd grow into that and use it over time. So my view on this is that rather than continuing to be constrained on CapEx ,and feeling in the core business like we have significant investments that we could make that we're not able to make that would be profitable, the right thing to do is to try to accelerate this to make sure that we have the compute that we need, both for the AI research and new things that we're doing and to try to get to a different state on our compute stance on the core business. Left totally unaddressed on the post-results analyst call - and perhaps an interesting reminder of the priorities of so many Wall Street watchers - was any probing about Meta's out-of-the-blue announcement last week of mass layoffs in its AI business. The loss of 600 positions across several areas of the AI operation was announced internally at Meta on 22nd October. In a memo to Meta staff, Chief AI Officer Alexandr Wang said that the cuts are intended to "streamline decision-making and increase the impact of individual roles". Meta previously announced 4,000 company-wide layoffs back in February. Zuckerberg said of the AI job losses last week that they do not reflect any change to Meta's commitment to AI - and his re-iteration in the face of Wall Street unhappiness about the need to keep on boosting CapEx here is at least him putting the company's money where his mouth is. Oh and as for that embarassing display last month when Zuckerberg ended up blaming his own firm's internet connection for a failed demo of the new AI glasses he's betting part of the farm on? Doesn't seem to have done much harm according to him:

3.5 Bn People Use At Least One Meta App Every Day: Zuckerberg in Q3 Earnings | AIM

However, the company's expenses also increased by 32% year over year, totalling $30.71 billion. Meta announced its financial results for the third quarter (Q3) of 2025 on 29 October. The company reported revenue of $51.24 billion, a 26% increase year-over-year. However, the company's expenses also increased by 32% year-over-year, totalling $30.71 billion. "We expect full year 2025 total expenses to be in the range of $116-118 billion, updated from our prior outlook of $114-118 billion and reflecting a growth rate of 22-24% year-over-year," said Susan Li, the CFO of the company. Additionally, Meta recorded a nearly $16 billion one-time charge for the 'One Big Beautiful Bill Act' introduced by US President Donald Trump. Without this fee, the company's net income would be approximately $18,64 billion, compared to the reported $2.71 billion. However, owing to this one-time charge, Meta said it expects a 'significant reduction in our U.S. federal cash tax payments for the remainder of 2025 and future years.' CEO Mark Zuckerberg described it as a 'strong quarter', revealing that 3.5 billion people are using at least one of Meta's apps each day. Instagram has now surpassed 3 billion monthly active users, and Threads has exceeded 150 million daily active users. "More than a billion monthly actives already use Meta AI, and we see usage increase as we improve our underlying models. I'm very excited to get a frontier model into Meta AI, and I think that the opportunity there is very large," Zuckerberg said. Over the past few months, Meta has made substantial investments in both talent and infrastructure to develop 'superintelligence' AI systems. This involves Meta purchasing a 49% stake in the AI startup Scale AI for $14.3 billion and appointing its CEO, Alexandr Wang, to head Meta's Superintelligence team. Besides, the company also reportedly shelled out lucrative pay packages exceeding $100 million to lure talent from other AI companies such as OpenAI, Google DeepMind and Anthropic. Recently, the company also announced its 25th data centre in the United States, in El Paso, Texas. Once completed, it will represent an investment of over $1.5 billion and will have a capacity of up to 1 GW. At a dinner gathering at the White House last month, Zuckerberg announced that the company would spend 'at least' $600 billion on US data centres and infrastructure through 2028. Regarding the timelines for developing 'superintelligent systems,' Zuckerberg suggested that instead of focusing on a specific number of years, "I think it's the right strategy to aggressively front-load building capacity so that way we're prepared for the most optimistic cases." "That way, if superintelligence arrives sooner, we will be ideally positioned for a generational paradigm shift and many large opportunities," he added. Zuckerberg said that if it takes longer, excess infrastructure will be used to accelerate the core business. "And in the worst case, we would just slow building new infrastructure for some period while we grow into what we build."

Meta's 2026 plan to spend even more on AI shakes investors

Mark Zuckerberg warned that Meta Platforms Inc. will need to spend even more aggressively on artificial intelligence in the year ahead, reigniting concerns from investors who sent its shares plunging on concerns that the massive investments won't pay off. In justifying the spending, the company's chief executive officer described an unsatiated appetite for more computing resources that Meta must work to fulfill to ensure it's a leader in a fast-moving AI race. "We want to make sure we're not underinvesting," he said on a call with analysts Wednesday after posting third-quarter results. Meta signaled in the earnings report that capital expenditures would be "notably larger" next year than in 2025, during which it expects to drop as much as $72 billion. Zuckerberg and Meta Chief Financial Officer Susan Li spent much of the call working to convince analysts that Meta's AI investments are paying off now -- and will in the future -- by helping the company better target ads and content. Revenue rose 26% to $51.2 billion in the third quarter, signaling that Meta's core advertising business, which generates about 98% of that revenue, remains strong. In the current period, Meta projected revenue will be $56 billion to $59 billion. Analysts, on average, estimated $57.4 billion, according to data compiled by Bloomberg. The Meta CEO wasn't concerned that the social media company may build more data centers and computing capacity than it needs to power its own AI systems. If Meta develops too much computing power, he said, the excess capacity could improve the core business. The company could also find a way to sell it, he said, the way rivals Microsoft Corp. and Alphabet Inc.'s Google already do. Zuckerberg's bigger point: "I think that it's the right strategy to aggressively front-load building capacity." In the face of Zuckerberg's comments, many Wall Street analysts were leery about the possibility of overspending. The shares of Menlo Park, California-based Meta slid as much as 14% on Thursday, the biggest intraday decline since April 2024. The plunge signals a shift in optimism from investors about the relentless spending: Meta shares had gained 28% this year through Wednesday's close. "Meta's earnings reveal the growing tension between the company's massive AI infrastructure investments and investor expectations for near-term returns, with rising spending on artificial intelligence capabilities weighing on sentiment despite solid underlying business performance," Jesse Cohen, a senior analyst at Investing.com, said in an email. That tension was evident on the earnings call, with one analyst saying to Zuckerberg that there were "obviously" some concerns on the investment. And yet, this year's spending spree isn't poised to stop. Meta said its capital expenditures for 2025 will be $70 billion to $72 billion, slightly raising the lower end of its previous outlook of $66 billion to $72 billion. So far this year, Meta has clocked $50 billion in capital expenditures, signaling more investment to come. Less than a day after issuing the projections, Meta was said to be targeting the sale of at least $25 billion in investment-grade bonds. The anticipated size, which Bloomberg reported Thursday according to people familiar with the matter, would make Meta's deal one of the largest of 2025. Li said Wednesday that Meta's total expenses will "grow at a significantly faster percentage rate in 2026 than 2025." She attributed the increasing costs to future infrastructure spending and the needs of Meta Superintelligence Labs. Zuckerberg commentary on spending has looked even further out. Speaking at a dinner President Donald Trump hosted for tech executives in September, the Facebook founder said that he expects the company to spend at least $600 billion in the US through 2028. Later that evening, Zuckerberg was caught on an open microphone telling Trump, "I wasn't sure which number you wanted to go with." Li shortly thereafter clarified at a conference that the $600 billion figure refers to the company's projections for total US spending this year through 2028, including on infrastructure, talent and other operations. Some analysts cautioned that the spending comments cast a shadow on what has otherwise been a thriving business. "Unfortunately, Meta's strong revenue and user growth in Q3 is tainted by significantly increased costs across the board," Mike Proulx, an analyst at Forrester, said in a note, also noting a streak of losses from Meta's AI-enabled hardware unit known as Reality Labs. Reality Labs reported a $4.4 billion operating loss in the third quarter, while bringing in $470 million in sales. The company also said that it expects lower year-over-year revenue from the unit in the fourth quarter, partially due to retail partners procuring its Quest virtual reality headsets during the current quarter to prepare for the holiday season. Li said the company expects sales of its AI-enabled glasses, such as the Ray-Ban Metas, to increase even as the Quest headsets face a slowdown. The company also reported third-quarter net income of $2.71 billion, which included a one-time, noncash income tax charge of $15.9 billion due to the implementation of the US tax bill signed into law in July, Meta said in the statement. Without the accounting charge, Meta said net income would have increased 19% to $18.6 billion. Looking beyond the third quarter, the company said it expects a "significant reduction" in US federal cash tax payments for 2025 and years to come due to the new law. -- With assistance from Kurt Wagner.

The AI bubble could pop the U.S. and global economies

Is artificial intelligence a bubble that is about to burst? If so, what would be the consequences for the U.S. and global economy? These issues have come to the fore as the U.S. stock market has surged to record highs on the back of massive gains by AI-related companies. Current valuations rival those during the peak of the tech boom in the second half of the 1990s. At the International Monetary Fund's annual meeting last week, the chief economist warned that the AI investment boom has "echoes" of the 1990s dot-com bubble. Previously, former IMF head Kristalina Georgieva said that indicators of financial instability were growing and investors should brace themselves for the challenges ahead. Countering this perspective, analysts from Goldman Sachs contend that "anticipated investment returns are sustainable." Their assessment calls for U.S. productivity to rise from 1.5 to 1.9 percent per year in the early 2030s. If so, they believe investments firms are undertaking to build data centers and expand software production should generate returns to justify current stock valuations. Nonetheless, while there is general agreement among experts that the adoption of AI has potential to transform the U.S. economy over time, some question whether the optimism about productivity gains will be borne out. Beyond this, there is uncertainty about the impact that a plunge in AI stocks might have on the U.S. and global economy. A key consideration is whether a tech selloff would mainly be felt by equity investors or would have broader ramifications. During the bursting of the dot.com bubble, for example, the stock selloff did not trigger a contraction of bank credit, as the Federal Reserve eased monetary policy considerably. As a result, the recession that ensued was relatively mild, with real GDP declining by about 0.3 percent. In comparison, the decline in home values in the mid-2000s rippled throughout the financial system and spawned a contraction in overall credit availability. It resulted in a more severe and protracted recession, in which real GDP fell by 4.2 percent from mid-2007 to 2009. The prevailing view until recently has been that the AI boom is mainly being financed by companies that have sizable cash holdings and strong balance sheets. If so, the risk of a selloff of AI stocks spreading throughout the economy could be mitigated. Those who are worried about a bubble, however, have expressed a variety of concerns. One is that a group of AI tech giants are funding each other through circular deals called "related party transactions." These deals, in which companies invest in or provide financing to customers, can inflate valuations by creating a distorted impression of growth prospects. Another is that some firms are increasingly turning to debt issuance to finance costly infrastructure such as data centers. The buildout of infrastructure has bolstered the economy this year, but it could add to risks if interest rates rise at some point. One of the International Monetary Fund's concerns relates to patchy oversight of nonbanks that include insurers, pension funds and hedge funds. This sector has grown rapidly and now holds roughly one half of the world's financial assets. So, what might this portend for the U.S. and global economy? Gita Gopinath, former chief economist of the IMF now with Harvard, calculates that a market correction of the same magnitude of the dot.com crash could wipe out about $20 trillion in wealth for American households. This would be equivalent to nearly 70 percent of U.S. GDP. She estimates that foreign investors could also face wealth losses exceeding $15 trillion, or about 20 percent of the rest of the world's GDP. Gopinath's conclusion is that the structural vulnerabilities and macroeconomic context are more perilous today than in the early 2000s, when one takes into account the fallout from higher tariffs, potential loss of Fed independence and loss of confidence in the U.S. dollar. My take is that an equity market selloff today may not be comparable to that of the early 2000s, when the S&P 500 and the NASDAQ Composite sold off by 50 percent and 75 percent, respectively. The reason is that, in addition to the overvaluation of tech stocks, two other forces kicked in then. One was the Sept. 11, 2001 attack, and the other was the Enron scandal in 2002, which caused investors to question how widespread accounting fraud was. That said, I share Gopinath's concerns about the current macroeconomic context. I also take note of the rapid expansion of margin debt, which reached a record high of $1.1 trillion in September and is up by 34 percent over a year ago according to the Financial Industry Regulatory Authority. The leverage on underlying positions can exacerbate losses if the stock market declines. Accordingly, a stock market selloff of 20 to 30 percent is possible, which could tip the U.S. economy into a moderate recession at some point. Nicholas Sargen, Ph.D., is an economic consultant for Fort Washington Investment Advisors and has authored "Global Shocks: An Investment Guide for Turbulent Markets."

We May Not Be in an AI Bubble, Reveals Report | AIM

Coatue Management, a US-based investment management firm, released a Public Markets Update report recently, addressing whether the AI market is in a bubble. The firm presents a bullish case, stating that AI represents essential infrastructure investment backed by $150 billion in current revenues and unprecedented adoption rates that justify capital expenditure surges. This contradicts the sentiments of market experts, who have cited multiple bubble indicators acknowledged by Coatue. The key arguments today are that the concentration of the top 10 companies matches the dot-com levels of 2000, and AI-related capital expenditure hit 1.5% of GDP and continues to rise. Besides, vendor financing appears circular. Companies like NVIDIA effectively fund their own future sales by investing in OpenAI, which commits spend to Oracle, which, in turn, buys NVIDIA hardware. Several reports claim AI adoption among enterprises is slow, with few tangible benefits or profits. However, Coatue counters these arguments with data. On capital expenditure, the firm noted that the top 10 tech companies generate approximately $1 trillion in annual free cash flow before capex. Unlike previous infrastructure booms requiring government funding or debt, this investment comes from private sector cash reserves. Philippe Laffont, founder of Coatue Management, linked AI investment to historically beneficial infrastructure. "These are some of the big infrastructure spends that we've had in the US, whether it's been the phone network, electrification, cars, the dams, the interstate highway, the internet, the personal computer, mobile internet....(sic)," he stated in a video explaining the report. The Nasdaq-100 next twelve months (NTM) price-to-earnings (P/E) multiple in 2025 stands at 28x, compared to 89x in 1999. This indicates a more reasonable valuation, rather than wild overpricing like the dot-com bubble. Besides, today's top tech companies trade at 28x forward earnings, down from 67x in 1999. This means while tech stocks are still considered expensive, investors are paying far less relative to the companies' profits than they did during the dot-com bubble -- when prices were driven more by hype than actual earnings. Source: Coatue Moreover, equity issuances (IPOs and secondary offerings) in 2025 stand at 56, sharply down from the 511 peak during the 2000 bubble, thereby suggesting that the companies are not rushing to go public recklessly. However, margin debt in US brokerage accounts reached 3.7% of the US GDP, near the 3.8% pandemic peak and close to the 3.0% level in 2000. This shows that individual investors are borrowing heavily, which could cause big losses if markets drop. Addressing the adoption concerns, Coatue stated that ChatGPT reached nearly one billion (~800 million) users faster than any technology in history, comparing favourably to Internet and PC adoption. Moreover, private AI startups with valuations exceeding $5 billion or $10 billion are rising. Even in terms of steady revenue, AI startups like Cursor have crossed $500 million in ARR, up from $300 million between mid-April and early June 2025. Source: Coatue While numerous reports state enterprises aren't yielding results from AI, several non-tech companies show real efficiency gains. As per the report, logistics company CH Robinson saw a 50% productivity improvement and approximately 30% headcount reduction, while Rocket Mortgage in financial services saved over six times more underwriting hours and achieved more than $40 million in annual run-rate cost savings. This proves that AI delivers tangible value for several non-tech companies as well. In addition, Coatue draws parallels to Microsoft's cloud business, which took six years to become profitable but now generates high returns. From a -7% return on investment capital (ROIC) in 2014, Azure reached 6% ROIC in 2018 and 33% in 2023. "Cloud didn't make money immediately, and today, the two most valuable businesses are Amazon and Microsoft," Laffont stated.

Meta's profit hit by $16 billion one-time tax charge

Meta faced a substantial charge impacting its stock. The company is heavily investing in AI infrastructure, planning massive data centers. This aggressive spending, shared by major tech firms, raises concerns about an AI bubble. Meta is also streamlining its AI unit with job cuts. These moves signal a focus on long-term AI development despite current cost pressures. Meta recorded a nearly $16 billion one-time charge in the third quarter related to U.S. President Donald Trump's Big Beautiful Bill, sending its shares down 8% after the bell. The social media company now expects capital expenditure to be between $70 billion to $72 billion, compared with its prior forecast of $66 billion to $72 billion. Meta continues to benefit from its massive user base. The company's powerful AI-optimized ad platform helps marketers automate campaigns, improve the quality of video ads, translate ads and generate persona-based images to target different customer segments. The company has launched ads on its messaging platform WhatsApp and social network Threads, directly competing with platforms such as Elon Musk's X, while Instagram's Reels continue to jostle with ByteDance's TikTok and YouTube Shorts for ad revenue in the short-video market. Meta has been doubling down on AI, with a target of achieving superintelligence, a theoretical milestone where machines could outthink humans. To that end, Meta reorganized its AI efforts under the Superintelligence Labs unit in June, following senior staff departures and a poor reception for its Llama 4 model. CEO Mark Zuckerberg has personally led an aggressive talent hiring spree and has said that the company would spend hundreds of billions of dollars to build several massive AI data centers for superintelligence. The company is among the top buyers of Nvidia's sought-after AI chips. The company struck a $27 billion financing deal last week with Blue Owl Capital, Meta's largest-ever private capital agreement, to fund a massive data center project in Richland Parish, Louisiana, known as "Hyperion." In a surprise move, Meta said last week it would cut around 600 jobs out of the several thousand employees within its AI unit to streamline decision-making and increase the responsibility, scope and impact of each role. The company's aggressive AI investments are creating significant cost pressures, even as it anticipates long-term benefits and revenue growth. Major tech companies including Alphabet, Amazon.com, Meta, Microsoft and CoreWeave are on track to spend $400 billion on AI infrastructure this year, Morgan Stanley estimates. These investments that come amid economic uncertainty have fueled fears of an AI bubble, putting pressure on CEOs to deliver measurable results, as the move could trigger losses, job cuts and boardroom shake-ups. (You can now subscribe to our ETMarkets WhatsApp channel)

Meta Flags Bigger 2026 AI Buildout as 'Superintelligence' Drive Accelerates | PYMNTS.com

By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. CFO Susan Lee said the company now expects capital-expenditure dollar growth to be "notably larger in 2026 than 2025," and total expenses to grow at a "significantly faster" rate next year. The biggest driver of those expenses will be compute -- both company‑owned data centers and third‑party cloud -- followed by compensation for artificial intelligence (AI) talent. Lee's written outlook in the earnings release echoed her message to analysts during the company's Q3 earnings call, noting expanding compute needs and "further upward pressure" on capital expenditures as Meta finalizes 2026 capacity plans. "I am very focused on establishing Meta as the leading frontier AI lab, building personal superintelligence for everyone," Zuckerberg told analysts, adding that the company is "aggressively front‑loading" capacity so it's prepared if milestones arrive sooner than expected. Zuckerberg framed the spend as both offense and insurance. If frontier breakthroughs take longer, the company will use additional compute to improve recommendations and ads, areas he said remain "compute‑starved." He also outlined how Meta plans to unify the "three giant transformers" that power Facebook, Instagram and ads into a single AI system to make trillions of daily recommendations across the apps. Meta reported Q3 revenue of $51.24 billion, up 26% year over year, with ad impressions up 14% and average price per ad up 10%. Family daily active people reached 3.54 billion, up 8%. A one‑time, non‑cash tax charge tied to U.S. tax law changes pushed the effective tax rate to 87% and reduced GAAP net income to $2.71 billion (EPS $1.05); excluding that charge, EPS would have been $7.25. On the product side, Instagram crossed 3 billion monthly actives, Threads continued to build momentum, and management reiterated plans to plug its next frontier model -- now being trained inside Meta Superintelligence Labs -- into Meta AI and business AI to deepen engagement and open new revenue streams.

Meta Q3FY25 Earnings Call - AI Drives Engagement and Adverts

During Meta's earnings call for 2025's third quarter (Q3FY25), Chief Executive Officer (CEO) Mark Zuckerberg revealed the multiple ways in which the tech giant is leveraging artificial intelligence (AI) across its verticals, and how the technology is driving engagement and revenue upwards. Notably, Zuckerberg remarked that AI-slop could be the third stage of the internet era after the user-driven and creator-driven content waves. "First was when all content was from friends, family, and accounts that you followed directly. The second was when we added all the creator content. Now, as AI makes it easier to create and remix content, we're going to add yet another huge corpus of content on top of those," he remarked. For context, Meta launched Vibes within the Meta AI chatbot app, which enables the users to create, remix, and share short-form AI-generated videos: highlighting the popular short-videos trend, which is happening across all major social media platforms. Meta claims that nearly 40% of the global population currently uses at least one app every day, saying that Instagram has 3 billion monthly active users and Threads, the recent addition to Meta's portfolio of social media apps, has passed 150 million daily active users. Meta revealed how the AI-driven recommendation system across its social media platforms is impacting engagement rates. Zuckerberg said that the AI recommendation system resulted in 5% and 10% more time spent on Facebook and Threads respectively in Q3FY25. Meanwhile, the watch time for video content, like reels on Instagram is much higher, standing at more than 30% since last year. As such, Zuckerberg remarked that Reels content has become very profitable for them. "As video continues to grow across our apps, Reels now has an annual run rate of over $50 billion," he said. Apart from the content recommendation systems, Meta is leveraging AI-enabled tools for delivering business services, including advertisements, as well. For context, Zuckerberg said that advertisers are spending over $60 billion per year using their end-to-end, fully AI-driven advertising systems. Elsewhere, Zuckerberg divulged that people have more than one billion active threads with business accounts across Meta's messaging platforms. Notably, he said that Meta's Business AI service will enable "tens of millions of businesses" to scale these conversations and improve sales at low cost. Furthermore, Zuckerberg said that Meta's consumer-facing Meta AI chatbot has "more than a billion" monthly active users. Importantly, he said that the tech giant is building a new unified AI system, which will combine the three major AI systems that run Facebook, Instagram, and ads recommendations. During the earnings call, Meta revealed how they are leveraging AI-based tools to encourage people to use AI-generated content as their advertisement material, including for creative output such as posters and music as well. Susan Li, Meta's Chief Financial Officer (CFO), while reporting on the company's financial performance, highlighted the growth in its Advantage+ creative suite. She said the number of advertisers using at least one video generation feature rose 20% from the previous quarter as adoption of image animation and video expansion tools increased. For context, Li was referring to Meta's Andromeda machine learning (ML) system for ad recommendations, which aims to significantly increase value for advertisers and users. She also revealed that the ad quality on Facebook has increased 14% because of the tech giant's AI-driven ad automation system. Emphasising "superintelligence for everyone", Zuckerberg said that he is "very focused on establishing Meta as the leading frontier AI lab". He also reiterated Meta's commitment to keeping AI innovation open source for others in the industry. Notably, this comes after a series of aggressive hirings, including poaching, for Meta's Superintelligence Labs, which was announced in June 2025 to build personal superintelligence that empowers everyone. Even during the previous quarterly earnings call, Zuckerberg referred to Superintelligence as their possible frontrunner in the future, hinting that AI superintelligence is a feasible possibility. Regarding the timeline of AI-powered superintelligence actually hitting the ground running, Zuckerberg said: "Some people think that we'll get there in a few years, others think it'll be five, seven years, or longer. I think it's the right strategy to aggressively front-load building capacity so that way we're prepared for the most optimistic cases. "That way, if superintelligence arrives sooner, we will be ideally positioned for a generational paradigm shift and many large opportunities." Importantly, he added that Meta plans to use additional computing capacity to aid their core businesses if superintelligence gets delayed. Meta's "Family of Apps", which includes Facebook, Instagram, WhatsApp, Messenger, Threads, Messenger, Business Suite, Meta AI, and Edits, reported the following key financial metrics in Q3FY25: Meanwhile, the US and Canada remain the largest revenue contributors, with the Asia-Pacific region following it. According to Meta's Earnings Presentation for this very quarter, the ad revenue for various regions is as follows: Notably, Meta continues to monitor ongoing legal and regulatory challenges in the US and the European Union (EU) that could significantly affect its business and financial results, Li said during the earnings call. In the EU, the company is working with the European Commission on its 'Less Personalised Ads' offering. Li warned that the Commission may impose changes that could hurt Meta's European revenue as early as the quarter under review. Meanwhile, in the US, several youth-related trials scheduled for 2026 could result in material losses.

Meta takes $16B hit to earnings from Trump's Big Beautiful Bill,...

Meta forecast "notably larger" capital expenses next year thanks to investments in AI, and recorded a nearly $16 billion one-time charge related to President Trump's 'Big Beautiful Bill' that pummeled its third-quarter profit. Shares of the company fell more than 6% after the bell. Excluding the charge, Meta said its third-quarter net income would have increased by $15.93 billion to $18.64 billion, compared to the reported net income of $2.71 billion. While third-quarter revenue beat estimates, the 26% growth was outpaced by a 33% increase in costs, which pressured margins. After a late start, Meta has doubled down on AI, with a target of achieving superintelligence, a theoretical milestone where machines could outthink humans. To that end, it has spent hundreds of billions of dollars to build several massive AI data centers for superintelligence and is planning for bigger financial outlays to meet big compute needs. "There's a range of timelines for when people think that we're going to get superintelligence," CEO Mark Zuckerberg said on a conference call with analysts. "I think that it's the right strategy to aggressively front load building capacity, so that way we're prepared for the most optimistic cases." If superintelligence takes longer than expected, then Meta will use the extra compute to accelerate its core business, and in the worst case, the company would slow building new infrastructure for some periods, he said. The social media company now expects capital expenditure between $70 billion and $72 billion this year, compared with its prior forecast of $66 billion to $72 billion. "We anticipate this will provide further upward pressure on our capital expenditures and expense plans next year." Employee compensation costs will be the second largest contributor to the increase in costs, Li said, to account for the compensation of employees hired throughout 2025, particularly AI talent. Meta continues to benefit from its massive user base. The company's powerful AI-optimized ad platform helps marketers automate campaigns, improve the quality of video ads, translate ads and generate persona-based images to target different customer segments. The company has launched ads on its messaging platform WhatsApp and social network Threads, directly competing with platforms such as Elon Musk's X, while Instagram's Reels continue to jostle with ByteDance's TikTok and YouTube Shorts for ad revenue in the short-video market. Meta has been doubling down on AI, with a target of achieving superintelligence, a theoretical milestone where machines could outthink humans. To that end, Meta reorganized its AI efforts under the Superintelligence Labs unit in June, following senior staff departures and a poor reception for its Llama 4 model. The company struck a $27 billion financing deal last week with Blue Owl Capital, Meta's largest-ever private capital agreement, to fund a massive data center project in Richland Parish, La., known as "Hyperion." In a surprise move, Meta said last week it would cut around 600 jobs out of the several thousand employees within its AI unit to streamline decision-making and increase the responsibility, scope and impact of each role. The company's aggressive AI investments are creating significant cost pressures, even as it anticipates long-term benefits and revenue growth. Major tech companies including Alphabet, Amazon, Meta, Microsoft and CoreWeave are on track to spend $400 billion on AI infrastructure this year, Morgan Stanley estimates. These investments that come amid economic uncertainty have fueled fears of an AI bubble, putting pressure on CEOs to deliver measurable results, as the push could trigger losses, job cuts and boardroom shake-ups.

Big Tech to report earnings under specter of AI bubble

As America's tech titans report earnings this week, one question looms large: is the artificial intelligence boom that has inflated valuations headed for the next big bubble? Microsoft, Alphabet, Amazon and Meta are poised to report that revenue rose at a brisk pace in the July-September quarter, according to LSEG data. The companies themselves are likely to say they will continue to pour billions into AI because it holds promise in the long term. But business leaders including OpenAI CEO Sam Altman, Amazon.com founder Jeff Bezos and Goldman Sachs GS.O CEO David Solomon have warned in recent months that the frenzy in tech stocks has outrun fundamentals. Investors, unnerved by the exuberance yet wary of betting against it, have started shifting away from hyped-up stocks, using dotcom-era strategies to dodge AI bubble risks. The four tech giants and other major cloud firms are together expected to spend $400 billion on AI infrastructure this year - but returns for businesses adopting the technology remain uncertain. A widely cited MIT study earlier this year found that of the more than 300 AI projects analyzed, only about 5 per cent delivered measurable gains. Most AI projects stall at the pilot stage due to weak integration into workflows and models that fail to scale, the study found. "Overall, the models are not there. I feel like the industry is making too big of a jump and is trying to pretend like this is amazing, and it's not. It's slop," OpenAI co-founder and Tesla's TSLA.O former AI head Andrej Karpathy said earlier this month. That could spell trouble for the AI-fueled rally that has added about $6 trillion to the Big Tech companies' market value since ChatGPT's November 2022 debut - and for the broader U.S. economy, which some economists say has been propped up by AI spending offsetting the drag from Trump-administration tariffs. Adding to the unease is a web of circular deals reminiscent of the 1990s dotcom boom, including Nvidia's potential $100 billion investment in OpenAI, one of its largest customers. OpenAI has signed AI compute deals worth $1 trillion with few details on how it will fund them, including a commitment to purchase $300 billion in computing power from Oracle ORCL.N. Debt is also playing a growing role in financing Big Tech's AI infrastructure spree in a departure from past investment cycles. Meta recently signed a $27 billion financing deal with private-credit firm Blue Owl Capital for its largest data center. "When the same companies are both funding and relying on each other, decisions may no longer be based on real demand or performance - but on reinforcing growth expectations," said Ahmed Banafa, engineering professor at San Jose State University. "These deals aren't necessarily problematic on their own - but when they become the norm, they increase systemic risk." Some investors said beneath the froth, real value is emerging - pointing to double-digit revenue growth and strong cash flows keeping Big Tech balance sheets healthy. "Adoption may be low right now but that's not a forward indicator. With greater spend and greater innovation in these models, the adoption is going to grow," said Eric Schiffer, CEO of Los Angeles-based investment firm Patriarch Organization, which holds shares in all the "Magnificent Seven" companies. "I don't think we are at a bubble stage yet." In the July-September quarter, the cloud-computing units of Amazon, Microsoft and Google are all expected to report strong growth despite capacity constraints limiting their ability to meet AI demand. They are also likely to reaffirm their capital spending plans. Microsoft Azure revenue likely rose 38.4 per cent in the period, outpacing expected growth of 30.1 per cent for Google Cloud and 18 per cent for Amazon Web Services, Visible Alpha data shows. AWS remains the largest player but has lagged Microsoft, which has benefited from its OpenAI tie-up, and Google, whose models have gained traction with startups. A recent AWS outage that disrupted several popular apps drew fresh scrutiny. Overall, Microsoft is expected to report revenue growth of 14.9 per cent in the quarter, while Alphabet's will likely rise 13.2 per cent, according to LSEG data. Amazon and Meta are likely to deliver revenue growth of 11.9 per cent and 21.7 per cent, respectively. Profit growth, however, is expected to slow for the companies as costs jump, with all barring Microsoft expected to post their weakest increase in 10 quarters. Microsoft, Alphabet and Meta will report results on Wednesday, followed by Amazon on Thursday.

Tech leaders ramp up AI spending, but Alphabet's cash flow wins investor favor

(Reuters) -Three of the biggest U.S. technology companies flagged plans on Wednesday to accelerate capital spending over the next year but investors were most accepting of Google-parent Alphabet's ability to fund its plans from its cash flow. Alphabet, Microsoft and Facebook-owner Meta all announced plans for higher annual capital expenditures as they pour money into chips and data centers. Shares of all three have risen substantially this year on expectations that they will be winners in the AI race, but investors only cheered Alphabet's report as they calculated the costs to each firm of the investments. All three reported stellar revenue growth in their key businesses, but investors pushed up Alphabet shares 6% and knocked down Microsoft by 4% and Meta by 8% in after-hours trading. A key reason for that split, analysts say, is Alphabet's ability to balance its soaring expenses with strong cash flow. "I would think that comes into play - to have capital spending be a lower percentage of revenue and cash flow. That maybe gives investors more comfort. All the players are ramping up spending pretty dramatically, and there's been a lot of concern about pressure on free cash flow," said Dave Heger, senior equity analyst Edward Jones. Alphabet's capital expenditure of $23.95 billion in the September quarter was 49% of its cash generated from operations. The percentage for Meta, however, is 64.6%, with Microsoft even higher at 77.5%. "Ongoing investments in data centers and AI infrastructure is a theme we've seen across Big Tech this earnings season. But unlike some of its peers, Alphabet is more than covering that spend with cash flow, and it's firing on all cylinders," said Josh Gilbert, market analyst at eToro. Investors have become wary of AI spending but big tech companies are not detailing exactly how much AI contributes to revenue and profit. With multi-billion-dollar deals being struck across the AI industry, investors are also growing cautious of a web of circular investments. Still, executives were adamant that they had to spend to keep up with demand for AI computing power. Meta CEO Mark Zuckerberg said that in the worst-case scenario of over-investing in AI, the company would see "some loss and depreciation, but we'd grow into that and use it over time." Companies with stronger cash flow can afford to invest more aggressively in AI infrastructure because they can tolerate lower returns on those outlays, said Dan Morgan, portfolio manager at Synovus Trust. Major cloud computing provider Amazon will offer another piece of the AI investment picture and returns in the tech space when it reports third-quarter earnings on Thursday. (Reporting by Deborah Sophia, Akash Sriram, Jaspreet Singh and Aditya Soni in Bengaluru; editing by Peter Henderson and Christian Schmollinger) By Deborah Mary Sophia, Akash Sriram and Jaspreet Singh

Meta forecasts bigger capital costs next year as Zuckerberg lays out aggressive AI buildout

(Reuters) -Meta on Wednesday forecast "notably larger" capital expenses next year thanks to investments in artificial intelligence, including aggressively building data centers to power its AI push. The Facebook and Instagram parent reported third-quarter revenue growth of 26% that beat market estimates, but that jump was outpaced by a 32% increase in costs. Shares of the company - that have risen 28% so far this year - fell 8% after the bell, as Wall Street digested Zuckerberg's even bigger capital plans to build out AI data center capacity that will pressure margins. Meta also recorded a nearly $16 billion one-time charge related to U.S. President Donald Trump's 'Big Beautiful Bill' that pummeled its third-quarter profit. Excluding the charge, net income in the quarter would have increased to $18.64 billion, compared with the reported net income of $2.71 billion. After a late start, Meta has doubled down on AI, with a target of achieving superintelligence, a theoretical milestone where machines outthink humans. To that end, it has pledged to spend hundreds of billions of dollars to build several massive AI data centers for superintelligence and is planning for bigger financial outlays to meet big compute needs. "There's a range of timelines for when people think that we're going to get superintelligence," CEO Mark Zuckerberg said on a conference call with analysts. "I think that it's the right strategy to aggressively front-load building capacity, so that way we're prepared for the most optimistic cases." If superintelligence takes longer than expected, then Meta will use the extra compute to accelerate its core business, and in the worst-case scenario, the company would slow building new infrastructure for some periods, he said. "After a few years of existential hand-wringing, the company has found its rhythm again by doing what it does best: scaling attention and monetizing it with ruthless efficiency," said Jeremy Goldman, a senior director at Emarketer. "While everyone else is still pitching AI moonshots, Meta has quietly turned AI into margin. Its ad tools are sharper, its targeting smarter, and its short-form video business is finally paying off." HIRING SPREE BOOSTS EXPENSE OUTLOOK Meta, determined to catch up with rivals including Microsoft and Alphabet's, has accelerated spending on AI, reorganizing its AI efforts under a "Superintelligence Labs" unit in June. Zuckerberg has personally led an aggressive spree to hire talent. The company is among the top buyers of Nvidia's sought-after AI chips. "Meta Superintelligence Labs is off to a strong start," Zuckerberg said. "I think that we've already built the lab with the highest talent density in the industry ... We're also building what we expect to be an industry-leading amount of compute." Employee compensation costs will be the second largest contributor to the increase in costs next year to account for the compensation of employees hired throughout 2025, particularly AI talent, Meta CFO Susan Li said in a statement. Between the three companies, plus Amazon and ChatGPT owner OpenAI, construction of AI datacenter capacity has skyrocketed. The costs to build these centers have fueled fears of an AI bubble, putting pressure on CEOs to deliver measurable results, and prompting them to find partners to help finance these outlays. On Wednesday, Alphabet and Microsoft also signaled higher AI investments. OpenAI CEO Sam Altman on Tuesday said he would eventually like the company to be able to add 1 gigawatt of compute every week - an astronomical sum given that each gigawatt currently comes with a capital cost of more than $40 billion. Meta also boosted the lower end of its capital expenditure outlook to between $70 billion and $72 billion this year, compared with its prior forecast of $66 billion to $72 billion. "Meta's earnings reveal the growing tension between the company's massive AI infrastructure investments and investor expectations for near-term returns, with rising spending on artificial intelligence capabilities weighing on sentiment despite solid underlying business performance," said Jesse Cohen, a senior analyst at Investing.com. MASSIVE USER BASE FUELS AD REVENUE Meta continues to benefit from its massive user base. The company's powerful AI-optimized ad platform helps marketers automate campaigns, improve the quality of video ads, translate ads and generate persona-based images to target different customer segments. The company estimates that more than 3.5 billion people used at least one of its apps on a daily basis last month. The company has launched ads on its messaging platform WhatsApp and social network Threads, directly competing with platforms such as Elon Musk's X, while Instagram's Reels continue to jostle with ByteDance's TikTok and YouTube Shorts for ad revenue in the short-video market. Meta said it expects fourth-quarter revenue between $56 billion and $59 billion, compared with analysts' average estimate of $57.25 billion, according to data compiled by LSEG. (Reporting by Jaspreet Singh in Bengaluru; Editing by Alan Barona, Deepa Babington and Sayantani Ghosh) By Jaspreet Singh and Echo Wang

Big Tech to report earnings under specter of AI bubble

(Reuters) -As America's tech titans report earnings this week, one question looms large: is the artificial intelligence boom that has inflated valuations headed for the next big bubble? Microsoft, Alphabet, Amazon and Meta are poised to report that revenue rose at a brisk pace in the July-September quarter, according to LSEG data. The companies themselves are likely to say they will continue to pour billions into AI because it holds promise in the long term. But business leaders including OpenAI CEO Sam Altman, Amazon.com founder Jeff Bezos and Goldman Sachs CEO David Solomon have warned in recent months that the frenzy in tech stocks has outrun fundamentals. Investors, unnerved by the exuberance yet wary of betting against it, have started shifting away from hyped-up stocks, using dotcom-era strategies to dodge AI bubble risks. AI RETURNS REMAIN UNCERTAIN The four tech giants and other major cloud firms are together expected to spend $400 billion on AI infrastructure this year - but returns for businesses adopting the technology remain uncertain. A widely cited MIT study earlier this year found that of the more than 300 AI projects analyzed, only about 5% delivered measurable gains. Most AI projects stall at the pilot stage due to weak integration into workflows and models that fail to scale, the study found. "Overall, the models are not there. I feel like the industry is making too big of a jump and is trying to pretend like this is amazing, and it's not. It's slop," OpenAI co-founder and Tesla's former AI head Andrej Karpathy said earlier this month. That could spell trouble for the AI-fueled rally that has added about $6 trillion to the Big Tech companies' market value since ChatGPT's November 2022 debut - and for the broader U.S. economy, which some economists say has been propped up by AI spending offsetting the drag from Trump-administration tariffs. CIRCULAR DEALS ADD TO THE NERVOUSNESS Adding to the unease is a web of circular deals reminiscent of the 1990s dotcom boom, including Nvidia's potential $100 billion investment in OpenAI, one of its largest customers. OpenAI has signed AI compute deals worth $1 trillion with few details on how it will fund them, including a commitment to purchase $300 billion in computing power from Oracle. Debt is also playing a growing role in financing Big Tech's AI infrastructure spree in a departure from past investment cycles. Meta recently signed a $27 billion financing deal with private-credit firm Blue Owl Capital for its largest data center. "When the same companies are both funding and relying on each other, decisions may no longer be based on real demand or performance - but on reinforcing growth expectations," said Ahmed Banafa, engineering professor at San Jose State University. "These deals aren't necessarily problematic on their own - but when they become the norm, they increase systemic risk." SOME INVESTORS BET ADOPTION WILL GROW Some investors said beneath the froth, real value is emerging - pointing to double-digit revenue growth and strong cash flows keeping Big Tech balance sheets healthy. "Adoption may be low right now but that's not a forward indicator. With greater spend and greater innovation in these models, the adoption is going to grow," said Eric Schiffer, CEO of Los Angeles-based investment firm Patriarch Organization, which holds shares in all the "Magnificent Seven" companies. "I don't think we are at a bubble stage yet." In the July-September quarter, the cloud-computing units of Amazon, Microsoft and Google are all expected to report strong growth despite capacity constraints limiting their ability to meet AI demand. They are also likely to reaffirm their capital spending plans. Microsoft Azure revenue likely rose 38.4% in the period, outpacing expected growth of 30.1% for Google Cloud and 18% for Amazon Web Services, Visible Alpha data shows. AWS remains the largest player but has lagged Microsoft, which has benefited from its OpenAI tie-up, and Google, whose models have gained traction with startups. A recent AWS outage that disrupted several popular apps drew fresh scrutiny. Overall, Microsoft is expected to report revenue growth of 14.9% in the quarter, while Alphabet's will likely rise 13.2%, according to LSEG data. Amazon and Meta are likely to deliver revenue growth of 11.9% and 21.7%, respectively. Profit growth, however, is expected to slow for the companies as costs jump, with all barring Microsoft expected to post their weakest increase in 10 quarters. Microsoft, Alphabet and Meta will report results on Wednesday, followed by Amazon on Thursday. (Reporting by Aditya Soni in Bengaluru; Editing by Sayantani Ghosh and Saumyadeb Chakrabarty)