Why Michael Burry Thinks the AI Boom Is a Bubble

Michael Burry, the legendary investor who famously bet against the U.S. subprime mortgage market and earned roughly $1 billion during the 2007–2008 financial crisis, is once again sounding the alarm.

In his Substack, Burry warns that the artificial intelligence frenzy has created a classic speculative bubble — one he believes is headed for a sharp reversal and a painful collapse. He argues that stock valuations have reached levels that are simply “too good to be true,” detached from economic reality.

Why Michael Burry Thinks the AI Boom Is a Bubble This is the same investor who has correctly anticipated major market turns before. Yet since 2023 he has issued roughly five separate warnings about an imminent stock-market crash — each time only to watch the market climb higher.

On January 31, 2023, when Burry first highlighted extreme overvaluation, the Nasdaq was up roughly 130 % by mid-2026.
By mid-August 2023, the index had still risen 93 % from his warning.
Even at the end of 2025, the Nasdaq stood 14 % higher than at the time of his call.
As recently as March 2026, it had gained another 18 %.

Burry is not merely venting frustration. He backs his thesis with concrete, uncomfortable arguments that even his critics have had to address.

1. “Creative” Accounting on AI Hardware

Why Michael Burry Thinks the AI Boom Is a Bubble Burry accuses Big Tech of systematically understating depreciation on the massive investments in AI chips and data-center servers. Companies are assuming these assets will last far longer than they realistically will.

According to Burry, the true useful life of today’s high-end AI hardware is closer to 2.5 years. If firms applied that realistic schedule, reported profits would look dramatically worse.

His estimates are striking:

Oracle is overstating profits by 48–62 %.
Meta and Microsoft are each inflating earnings by more than 20 %.

2. Enron-Style Financing and the Debt Spiral

Why Michael Burry Thinks the AI Boom Is a Bubble Burry draws an unflattering parallel to Enron’s off-balance-sheet schemes. AI-heavy companies are borrowing enormous sums, using rapidly depreciating chips as collateral.

Because the equipment loses value so quickly, firms must constantly raise fresh debt just to buy the next generation of hardware. The result, he says, is a self-reinforcing debt spiral that masks true financial fragility.

3. The Dot-Com Parallel: Hype vs. Reality

The investor explicitly compares today’s AI mania to the dot-com bubble of 1999–2000. In both cases, sky-high market valuations have little to do with current earnings or sustainable cash flows. Burry calls the impending collapse “predictable,” pointing to March 2000 as the moment when the music finally stopped for internet stocks. He believes the same reckoning is coming for AI.

Burry’s Actual Portfolio: Shorts and Longs

Burry does not simply talk — he puts capital behind his convictions. His latest 13F filing reveals a carefully constructed book that is simultaneously short the AI hype and long the parts of the market the AI frenzy has unfairly ignored.

Why Michael Burry Thinks the AI Boom Is a Bubble Short positions (bearish bets):

Palantir (PLTR) — his largest single position: $912.1 million in put options.
NVIDIA (NVDA) — $186.6 million in put options, targeting the undisputed king of AI chips.
Oracle (ORCL), plus broad hedges via QQQ (Nasdaq-100 ETF) and SOXX (semiconductor ETF).

Long positions (bullish bets):

Traditional enterprise software companies that were punished by the AI narrative despite strong fundamentals: Adobe, Autodesk, Salesforce, and Veeva Systems.

In short, Burry is short the “inflated AI dreams” and long the “undervalued reality.”

The Overlooked Energy Dimension

Why Michael Burry Thinks the AI Boom Is a Bubble One of Burry’s most sobering points extends beyond Wall Street. The AI boom has made the United States dangerously dependent on energy-hungry data centers. These facilities already consume about 4 % of all U.S. electricity, and the figure is rising fast.

Should the AI market correct sharply, demand for chips, servers, and the electricity that powers them would collapse together. The fallout would not be confined to tech stocks — it would ripple into the real economy, hitting power generation, construction, and related supply chains.

What This Means for Investors

Burry’s warning is not investment advice; it is a cautionary signal. Even if he is ultimately right about the bubble, history shows that bubbles can inflate far beyond what seems rational. He himself acknowledges that the final crash may still be a year or two away, and the market could climb further in the meantime.

For anyone with exposure to AI-related stocks, the message is clear:

Re-examine your portfolio for concentration risk.
Consider diversification beyond the current hype cycle.
Maintain a sober perspective when everyone else is euphoric.

Michael Burry’s record proves that being early can be painful — but being right eventually matters more. Whether the AI bubble bursts in 2026, 2027, or later, his analysis forces a necessary question: are today’s valuations built on sustainable value, or on the same collective delusion that has ended every previous tech mania?