Michael Burry’s Latest Warning on Amazon, Meta and Alphabet AI Spending

TLDR

• Michael Burry says Amazon, Meta, Alphabet, Microsoft and Oracle carry roughly $3 trillion in AI-related spending commitments.
• He compares the current AI boom to the dot-com bubble and expects write-offs around 2028 or 2029.
• Nvidia sent a seven-page memo disputing Burry’s claims about chip depreciation timelines.
• Burry has increased short positions on Micron, Nebius, Palantir and semiconductor stocks.
• He is also buying discounted stocks like Build-A-Bear, Birkenstock and Sprouts Farmers Market.

---

Michael Burry has issued a fresh warning about Big Tech spending on artificial intelligence. The investor, known for predicting the 2008 housing crash, says the spending pattern looks like past bubbles that ended badly.

🚨 Michael Burry says AI companies are on track to outspend even the dot-com bubble.

He says overspending like this has marked the top of every major bubble over the past 40 years.

S&P 500 net investment hit 2.07% of GDP in Q2, the highest level in almost 40 years except the… pic.twitter.com/3NXsy4SW1i

— Bull Theory (@BullTheoryio) September 27, 2026

Burry made his case in a Substack post published on September 24. He focused on five companies: Amazon, Meta, Alphabet, Microsoft and Oracle.

Burry’s $3 Trillion Estimate

Burry says these companies together carry close to $3 trillion in spending commitments tied to AI infrastructure. This includes purchase obligations, future leases, guarantees and construction costs.

Alphabet drew the most attention in his post. Burry estimates the company holds nearly $900 billion in off-balance-sheet commitments tied to its AI buildout.

Meta’s uncommenced leases and purchase obligations come to roughly $700 billion, according to Burry. He said that number could approach $1 trillion once broader commitments are included.

Burry compared the current spending boom to the dot-com era. He pointed to net capital investment across S&P 500 companies reaching its highest share of GDP in about four decades.

He does not expect problems to show up right away. Instead, he believes write-offs could emerge around 2028 or 2029.

Burry has also raised concerns about how companies account for their Nvidia chips. He argues hyperscalers are stretching the useful life of these chips beyond a realistic two to three year cycle.

That accounting choice, in his view, could understate industry-wide depreciation by about $176 billion between 2026 and 2028.

Nvidia and Micron Push Back

Nvidia responded directly to Burry’s claims. The company sent a seven-page memo to Wall Street analysts arguing that four to six years is a more accurate depreciation window for its chips.

Nvidia also disputed a specific figure from Burry, saying it repurchased $91 billion in shares since 2018, not the $112.5 billion he cited.

Micron’s chief business officer told investors that demand for memory chips remains above what the company can supply through 2028. That statement runs counter to Burry’s view that current AI demand is overstated.

Burry has built short positions tied to his thesis. These include bets against Oracle, Nebius, Micron and Palantir.

He has also shorted the Philadelphia Semiconductor Index through January 2027 put options. Most of the stocks he is shorting posted gains through August.

Not everyone dismisses Burry’s concerns. GMO co-founder Jeremy Grantham has called current AI valuations a bubble. DoubleLine Capital’s Jeffrey Gundlach has said he expects losers to emerge from the AI race.

Burry Is Buying Some Stocks Too

Burry is not betting against the entire market. He has taken new full positions in five companies outside the AI sector.

These include roofing distributor QXO, grocery chain Sprouts Farmers Market, toy retailer Build-A-Bear, footwear brand Birkenstock and e-commerce company Mercado Libre.

All five stocks have fallen double digits this year. Build-A-Bear is down as much as 57%.

Micron is set to report quarterly earnings soon. Wall Street expects year-over-year earnings growth of 940%, according to Investor’s Business Daily. That report could shape how investors view the broader AI spending debate going forward.