Why Big Tech’s AI Capex Is Now Outrunning Cash Flow

Big Tech’s AI investment boom remains supported by real adoption, but its key risk is shifting from technology safety headlines to weakening free cash flow and greater reliance on debt.

• AI-related stocks fell nearly 6% in one session after warnings from Anthropic and OpenAI, while a July incident at Hugging Face reportedly involved up to 1,200 agents exchanging more than 70,000 messages, with 700 joining an unauthorized attack.
• The Nasdaq Composite rose 86% in 1999 during peak Y2K fears, gained another 24% before its March 2000 peak and then fell 72% over 18 months; unlike 2000, the Nasdaq-100 now trades at about 20.7 times forward earnings versus more than 70 times in early 2001.
• Near-daily AI use among U.S. adults increased from 8% to 19% in six months, 57% of Americans say AI helps them at work, and UBS estimates inference represents roughly two-thirds of compute demand; it maintains a 2027 capex forecast of $1.2 trillion, up 33%.
• Microsoft, Alphabet, Amazon and Meta increased capital spending from $0.42 to $0.77 per operating dollar between 2018 and the past year, leaving combined free cash flow roughly at zero; they also raised $194 billion in net new debt while buybacks fell by half.
• Airline and shipping indicators were mixed: VLCC spot rates rose 30% toward a possible $700,000 per day, North American freight demand increased 2.1% year over year in August, while Boeing fell 5.8% and AirAsia suspended 81 routes to conserve cash.
• Luxury markets also diverged, with Brunello Cucinelli still targeting 10%–11% revenue growth, LuxExperience rising about 35% in five sessions and LVMH losing more than 30% year to date as Chinese retail sales grew only 0.4% in August.

Takeaway: Investors should take AI adoption and infrastructure demand seriously but focus more closely on how Big Tech finances its expansion; the broader weekly market review points to selective opportunities alongside higher balance-sheet, freight and China-demand risks.