Leopold Aschenbrenner’s AI fund unraveled just before the tech-stock rebound

Leopold Aschenbrenner’s AI fund unraveled just before the tech-stock rebound

N
News Editor
2026-07-31 08:33:48
Former OpenAI researcher Leopold Aschenbrenner, once hailed on Wall Street as a rising AI investing star, saw his Situational Awareness fund collapse on July 30 after a concentrated, leveraged bet on AI hardware and large-cap technology stocks turned against him. The fund, which reportedly swelled to about $45 billion at its peak, had built a strategy around going long AI infrastructure names and shorting traditional software. That trade came under pressure as core holdings fell, short positions rebounded, and prime brokers including Bank of America, Goldman Sachs, and JPMorgan began issuing margin calls. Part of the fund’s publicly traded portfolio financed with borrowed money was acquired by Citadel Investment Group, the sister company of Citadel Securities, while the portion backed by client capital remained in place, according to people familiar with the matter. After the transaction, total assets reportedly dropped from roughly $45 billion in early July to about $10 billion. The blowup landed as investors were reassessing a broader theme behind the AI trade: major technology companies are still delivering strong cloud and revenue growth, but heavy spending on data centers and compute infrastructure is putting free cash flow under strain. Recent results from Alphabet, Meta, and Amazon sharpened that concern, even as tech stocks staged a sharp rebound on July 30.

US technology stocks reversed sharply on July 30. After broad weakness the day before, the Nasdaq rose 2.8%, while Microsoft jumped 15.51% in its biggest one-day gain since 2008.

One day earlier, former OpenAI researcher Leopold Aschenbrenner had already been forced out of the market. Aschenbrenner, widely cast on Wall Street as a new AI investing prodigy, saw his Situational Awareness fund blow up just before the rebound.

Only two months earlier, the fund had reached near-mythic status in the market. Its assets reportedly swelled to about $45 billion, with exposure heavily concentrated in AI and large technology stocks. Less than 10 days later, the backdrop changed abruptly.

Alphabet and Tesla both reported negative free cash flow. The Magnificent Seven index then fell 4.8% in a single session, wiping about $797 billion from the group’s combined market value. Meta followed with a 91% drop in second-quarter free cash flow to just $784 million. Citadel Securities added to the pressure in a research note that said the Federal Reserve could unexpectedly raise rates, darkening sentiment further.

Higher long-end yields, runaway AI capital spending, worsening cash-flow profiles at technology giants, and a concentrated leveraged book left little room to maneuver. According to market accounts, Aschenbrenner sold most of his public-equity portfolio to Citadel Investment Group, the sister company of Citadel Securities.

A 25-year-old former OpenAI researcher built a four-times leveraged AI trade

Citadel Investment Group, which took over the Situational Awareness assets, was founded by billionaire Ken Griffin and is one of the world’s largest hedge funds.

Aschenbrenner is 25 and was born in Germany. He entered Columbia University at 15 and graduated at 19 as a student commencement speaker. Economist Tyler Cowen described him as an “economics wunderkind.”

In 2024, OpenAI fired Aschenbrenner over what the company described as an improper disclosure of internal information. After leaving, he published a 165-page essay in June of that year titled Situational Awareness: The Decade Ahead. The piece circulated widely in Silicon Valley and later became the foundation for his fundraising.

He launched Situational Awareness in September 2024 with only a few hundred million dollars in initial capital.

The strategy was tightly concentrated: long AI hardware and infrastructure, short traditional software. Regulatory filings show that by the end of the first quarter of 2026, the fund’s largest public-market positions included Nebius, SanDisk, Micron, and CoreWeave. It also held major stakes in Bloom Energy, SK Hynix, Oracle, and AMD, while shorting software companies including Adobe.

Situational Awareness used borrowing to amplify returns, and leverage at one point reached four times. In private markets, it held stakes in Anthropic, chip startup MatX, and data-center company Fluidstack.

By the end of 2025, the firm had about eight employees, only four of whom were investment professionals.

The trade started to break on July 10 and unraveled within days

The turning point came on July 10.

That day, SK Hynix American depositary receipts began trading on Nasdaq at an offering price of $149 per receipt. The deal raised about $26.5 billion, making it the largest US initial public offering by a foreign company. Situational Awareness was one of the cornerstone investors in the listing.

On its first trading day, the SK Hynix ADR opened at $170, up about 14%. The strength did not last. AI hardware names soon began to retreat as a group.

From July 10 to July 20, pressure built on both sides of the fund’s book. On the long side, key holdings including SK Hynix, SanDisk, Bloom Energy, and Nebius kept falling, with those stocks generally down more than 30% over one month. On the short side, software shares including Adobe rebounded, hurting the fund’s long-hardware, short-software trade at both ends.

On July 24, Aschenbrenner wrote to investors saying the fund had “not been spared” by the market turmoil, particularly in Asia. He still called the selloff one of the most attractive buying opportunities since early 2025 and invited clients to commit new capital starting Aug. 1. Bloomberg reported that the appeal did not produce the funding commitments he had hoped for.

Three days later, Citadel Securities publicly said the Fed could unexpectedly raise rates, adding to the downturn in sentiment. By July 29, prime brokers including Bank of America, Goldman Sachs, and JPMorgan had begun issuing margin calls.

On July 30, Situational Awareness formally collapsed.

People familiar with the matter said Citadel Investment Group acquired the part of the public-equity portfolio financed with borrowed money, while the remaining portion funded by clients was retained. After the deal, total assets at Situational Awareness reportedly dropped from a peak of about $45 billion in early July to roughly $10 billion.

The timing added another striking detail. The blowup came in the same week as Aschenbrenner’s upcoming wedding. His fiancée, Avital Balwit, serves as chief of staff to Anthropic CEO Dario Amodei.

Big Tech revenue stayed strong, but AI spending hit free cash flow

The collapse of Situational Awareness also reflected a broader turn in the AI trade. Technology giants once valued for fortress-like cash generation are now pouring cash into data centers and compute infrastructure.

For years, investors were willing to support elevated valuations partly because these companies combined growth with large and durable free cash flow. Recent earnings suggested that assumption is weakening.

Alphabet

Alphabet reported second-quarter revenue of $119.8 billion, up 24% year over year. Google Cloud revenue reached $24.8 billion, up 82%, and operating margin in the cloud unit rose from 20.7% to 35.6%. At the same time, capital expenditures hit $44.9 billion, pushing quarterly free cash flow to negative $5.9 billion.

Chief Financial Officer Anat Ashkenazi said on the earnings call that free cash flow “will continue to be pressured” as AI investment deepens.

Meta

Meta showed a similar pattern. Second-quarter revenue came in at $60.8 billion, up 28%, and advertising revenue continued to grow. But costs and expenses jumped 55% from a year earlier, capital expenditures reached $31.08 billion, and free cash flow fell 91% to $784 million. Management said expanding compute capacity remains a priority through 2026 and 2027.

Amazon

Amazon reported second-quarter AWS revenue of $42.2 billion, up 37%, its fastest growth rate in 18 quarters. The company also raised its full-year capital spending forecast to $220 billion. Over the trailing 12 months, free cash flow swung from a net inflow of $18.2 billion in the prior-year period to a net outflow of $7.6 billion.

Combining full-year guidance for the largest spenders produces a staggering figure: $205 billion for Alphabet, $220 billion for Amazon, $190 billion for Microsoft, and $145 billion for Meta, or about $760 billion in total.

“The real issue is the scale of spending right now. Nobody knows what the return on investment is,” said Ken Mahoney, chief executive of Mahoney Asset Management.

The biggest seller may be gone, but the AI trade has not fully cleared

The July 30 rebound in tech stocks was viewed in part as relief after the market’s biggest forced seller had exited.

O’Rourke said that once investors know a potential seller is out of the way, dip buyers tend to rush in. CNBC commentator Jim Cramer called the liquidation at Situational Awareness a “clearing event” and said it could help stabilize the AI trade.

Even so, opinions on what the blowup means are not aligned.

Michael Dempsey, managing partner at Compound, wrote that Situational Awareness had “incredibly impressive foresight,” but said anyone with public-market experience could see it was a “one-way bet” with flawed risk management.

Former Bloomberg columnist Tim Culpan said the episode “shows the difference between wisdom and intelligence,” arguing that even a correct long-term thesis can fail when leverage removes the ability to withstand a sharp drawdown.

Investor Martin Shkreli put it more bluntly: with leverage at that level, there was no other ending.

Situational Awareness is not alone. Goldman Sachs prime brokerage data showed that Asia fundamental long-short funds were down 18.6% on average in July. A Morgan Stanley prime brokerage report said equity hedge funds were cutting both longs and shorts in roughly equal size, a sign that deleveraging is still underway.

The unresolved questions now are how long investors will tolerate strong cloud revenue paired with persistently negative free cash flow, and how many institutions still hold similarly concentrated, highly leveraged positions in the same stocks.

This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan.
600

Disclaimer:

The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.

Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.