Michael Burry's Bearish Stance on the Generative AI Market
Published on 29.11.2025
Going Short on Generative AI
TLDR: Legendary investor Michael Burry is now publicly bearish on the Generative AI market, citing flattening adoption rates, questionable ROI on massive infrastructure spending, and accounting tricks that overstate profits. He compares the current hype to previous bubbles, such as the dot-com era, and has taken significant short positions against key players like Nvidia.
Summary: Michael Burry, the investor famed for predicting the 2008 housing crisis, has turned his attention to the Generative AI boom, and his prognosis is anything but optimistic. He argues that despite the widespread hype, the economic impact of generative AI remains minimal, and adoption is plateauing. Data from the Census Bureau and a Brookings survey support this, indicating that most Americans use Gen AI only casually, with usage struggling to gain traction in professional settings, particularly among those without higher education. Burry points to a significant disconnect between the explosive capital expenditures by hyperscalers—projected at $3 trillion over the next five years—and the actual, limited end-user demand.
A central pillar of Burry's thesis is his critique of accounting practices, where companies are allegedly extending the depreciation timelines for AI hardware to inflate short-term profits. He asserts that this financial engineering masks the rapid obsolescence of technology, like Nvidia's GPUs, and creates a distorted view of profitability. For a company like Microsoft, he estimates this could lead to an annual overstatement of $10-12 billion in net income. This situation, he suggests, is unsustainable and will likely result in future writedowns.
Burry draws parallels to the dot-com bubble, highlighting the speculative excess and "blessed fraud" that characterize late-stage manias. He believes Nvidia is particularly vulnerable, as its revenue is almost entirely dependent on the AI infrastructure build-out. Any pullback in spending by its major clients could trigger a sharp decline. He has disclosed bearish put options on both Nvidia and Palantir, signaling a strong conviction in his analysis. The debate is further fueled by Nvidia's own defensive memo, which Burry dismissed as a series of "straw man arguments," reinforcing his view that the market is detached from fundamentals.
Key takeaways:
- Generative AI adoption is flattening, and its use in professional settings is not as widespread as the hype suggests.
- The massive capital expenditure on AI infrastructure may not be justified by the current level of end-user demand and ROI.
- Accounting practices, such as extending hardware depreciation, could be artificially inflating the profits of major AI players.
- The current AI market shows signs of a speculative bubble, with parallels to past financial manias.
Tradeoffs:
- Investing heavily in AI infrastructure may position companies for future breakthroughs, but it comes at the risk of massive writedowns if demand does not materialize as projected.
- Optimistic accounting can boost short-term stock performance but sacrifices long-term financial transparency and investor trust.