Skip to content
NewSQL, Python and Power BI training for your team
Log in

The AI Bill Comes Due: Why Investors Balked at Big Tech’s Spending

Alphabet, Amazon and Intel lost hundreds of billions in one week. Here is what capex and free cash flow reveal.

Muhammad Jamilu Abdullahi 6 min read
Cover: a crumpled Alphabet Q3 receipt stamped "Payment due" showing free cash flow of minus $9.1bn, beside a falling stock chart and the headline "The AI Bill Comes Due"

A Wednesday evening shock

Shortly after Wall Street's closing bell on Wednesday 28 October, late evening in Lagos, Alphabet released its third-quarter results. At first glance, Google's parent company had a great story to tell. Google Cloud revenue had more than doubled, to $30.7bn.

Then investors reached the cash numbers. Alphabet had spent a record $51.4bn on capital expenditure in three months, against $24bn a year earlier. Its free cash flow was minus $9.1bn.

The shares fell more than 11% in after-hours trading. By Thursday's close they were down 12.6%, the steepest one-day fall since Google's 2004 stock market debut. Nearly half a trillion dollars of market value disappeared in a single session.

What happened, company by company

Over three days, the market judged each tech giant on one question: when does the AI spending start paying for itself?

CompanyKey numberThe worryShare reaction
AlphabetCloud revenue $30.7bn, up 103%Q3 capex $51.4bn; free cash flow −$9.1bn−12.6% on Thu 29 Oct
AmazonAWS revenue $46bn, up 39%Q3 capex about $76bn; 12-month free cash flow −$42.4bnAbout −2% on Fri, after falling as much as 6%
MicrosoftAzure growth of 48%Capex $45.2bn, below forecasts; free cash flow +$18.6bnRoughly flat (−0.2%) on Thu
IntelQ4 revenue forecast $15bn–$16bnWeak outlook; cash burn this yearRoughly −29% to −30% on Fri 30 Oct

Amazon's cloud arm, AWS, grew at its fastest pace in years. Yet Amazon missed Wall Street's forecasts for both revenue and profit. Its free cash flow over the past year swung from plus $14.8bn to minus $42.4bn.

Meta also reported negative free cash flow, of $2.1bn. Microsoft was the only one of the four big cloud spenders with cash left over after investment.

Intel delivered the week's harshest blow on Friday 30 October. It warned of a weak final quarter and of burning cash this year, citing soft PC demand, factory hurdles and heavy expansion spending. Its shares had their worst day since 1974.

That Friday, the Nasdaq Composite fell nearly 3%. Nasdaq 100 companies lost more than $1 trillion of market value in one session. October became the Nasdaq's worst month since March 2025.

Capex and free cash flow, in plain words

Capital expenditure, or capex, is money spent on assets a business will use for years. For Big Tech, that means data centres, AI chips, servers and networking gear. For a Kano bakery, it could be a new oven or a generator.

Free cash flow is the cash left after running the business and paying for those assets. The basic formula is operating cash flow minus capex. It is the money available for dividends, share buybacks, paying down debt or saving.

Alphabet's quarter makes the formula concrete. Its operations brought in $42.3bn of cash, but capex was $51.4bn. Subtract one from the other and you get minus $9.1bn.

Finance learners should notice something else. Alphabet's operating income still rose 30%, to $40.6bn. Profit and cash differ because capex is not charged all at once; it is spread over the asset's life as depreciation.

Alphabet said a build-up of its own TPU chip inventory and a $5.2bn European Commission fine payment also drained cash. Even so, Fortune counted four of the five biggest AI spenders with negative free cash flow in the quarter.

Fortune tallies up to $828bn of 2026 capex across Amazon, Alphabet, Microsoft, Meta and SpaceX, which now includes xAI. Amazon alone expects about $220bn; Alphabet guides $195bn–$205bn.

Why should anyone outside Wall Street care? When cash runs short, companies borrow or raise new money. Even cash-rich Microsoft launched a $25bn bond sale, its first in more than nine years.

Economist Aleksandar Tomic told Fortune that heavier tech borrowing could raise the cost of capital for other businesses. The AI bill can spill over to firms that never bought an AI chip.

The other side: bulls and bears

This is a summary of both camps, not investment advice. The bulls point to demand: Google Cloud has a $561bn backlog of contracted business. Finance chief Anat Ashkenazi said Alphabet expects returns to stay well above its cost of capital.

Amazon chief executive Andy Jassy called the AI build-out “a once-in-a-lifetime opportunity”. He expects AWS to grow by more than 40% in both 2027 and 2028. Microsoft's finance chief, Amy Hood, said: “We still have demand that exceeds supply.”

Some analysts saw an overreaction. CFRA's Angelo Zino suggested Alphabet's fall was probably a buying opportunity. B. Riley Wealth's Art Hogan described Friday as a rotation out of a crowded trade, not a broad reset.

The bears focus on cash, and on who benefits. Analyst Luke Lango, quoted by Fortune, argued that Alphabet is now funding AI with shareholders' money instead of its own cash. Capital.com's Daniela Hathorn said “markets are no longer rewarding companies simply for investing heavily in AI”.

The sharpest critique came days earlier. Apollo's chief economist, Torsten Slok, wrote that AI has delivered “zero financial payoff to corporate America outside of tech”. He found revenue per worker and margins rising at the seven tech giants but flat across the other 493 firms in the S&P 500.

What it means for Nigerian businesses and learners

The same question is arriving closer to home. On 28 October, BusinessDay reported that African AI startups are being squeezed by computing bills priced in dollars, while customers resist higher prices.

Local capacity is growing. In May, Kasi Cloud commissioned an AI-capable data centre campus in Lekki, Lagos. The company estimates Nigerian enterprises spend about $850m a year on foreign cloud infrastructure.

What follows is practical advice, not reported fact:

  • Treat AI tools as investments. Before subscribing, write down the cost, the hours the tool should save and the revenue it could add. Review the numbers after 90 days.
  • Watch your own free cash flow. A firm that is profitable on paper can still run short of cash after prepaying big annual software bills.
  • Mind the currency. Dollar-priced subscriptions get dearer in naira whenever the exchange rate moves against you. Track them monthly.
  • Pilot before you scale. Test one tool with one team, then expand.
  • Learners, read the cash flow statement. Building a simple Excel tracker of cost, time saved and payback period is a skill worth practising.

What to watch next

  • Alphabet's 2027 plan. It expects capex to rise significantly next year and will give figures on its fourth-quarter call.
  • The $1 trillion mark. Evercore and Bank of America expect AI-related capex to reach $1 trillion in 2027, according to Fortune.
  • Interest rates. Traders now expect US rate rises of more than half a percentage point by mid-2027, adding to worries about debt-funded AI spending.
  • The “S&P 493”. If ordinary companies start showing AI gains in revenue or margins, the bears' main argument weakens.

The AI bill has arrived. The next year will show who pays it, and who gets paid.

Sources

Keep reading

All posts