Michael Burry, the investor known for predicting the 2008 housing crash, has sharpened his bearish positioning against the AI boom, shifting several direct stock shorts into put options and warning that the market’s artificial intelligence bubble could burst sooner than he previously expected.
In his latest portfolio update, Burry said he covered short positions in Nvidia, Palantir, Micron, Nebius and other AI-linked names while moving much of that exposure into puts. He said the shift does not mean he is abandoning his bearish view, but that options provide more capital-efficient leverage over the shorter time horizon he now anticipates.
Burry’s Micron exposure moved into June puts with strikes near $500, while Nvidia was shifted into September 2027 contracts with strikes in the mid-$100s. He also increased his Palantir put exposure, concentrating positions around September 2027 strikes in the low $100s. Nebius, Oracle and semiconductor exposure through SOXX were also included in the repositioning.
Burry wrote that he was “moving timelines up,” arguing that the bubble may burst “sooner than later.” He had previously pointed toward 2028 as a potential period when excess computing capacity could become more obvious, but his latest trades suggest he now sees the possibility of market stress considerably earlier.
His argument centers on whether current valuations and infrastructure spending can be sustained long enough to justify investor expectations. Burry has questioned whether massive capital spending on data centers and chips will produce steady long-term returns, and he has warned that concentrated positioning and leverage could trigger a much sharper stock-market reversal if sentiment turns.
In a separate Substack post, Burry said the Trump administration sees the AI buildout as the main force holding up the U.S. economy and has little room to let AI investment slow down. “They cannot afford to let it fall,” Burry wrote. President Trump has said he will not slow AI development, describing the industry as potentially worth trillions of dollars and arguing that keeping U.S. leadership in AI matters more than managing potential risks.
Trump’s July trading disclosures showed more than 1,000 trades, mostly involving technology companies. He sold between $5 million and $25 million each in Microsoft and Amazon, then bought back smaller amounts shortly after, and also traded Oracle, Intuit, Marvell, Salesforce, Meta Platforms, Nvidia, ServiceNow, Workday and Adobe.
Burry’s positioning targets some of the biggest winners of the AI investment cycle. Micron has surged more than 270% this year, Nvidia remains the dominant supplier of AI accelerators and one of the largest companies in the S&P 500, and Palantir has become one of the market’s most prominent AI-linked stocks. AI names now represent an unusually large share of the S&P 500’s biggest holdings, making the broader market increasingly sensitive to changes in AI valuations.
