"This month has disappointed you": 25-year-old AI genius's confession of a $20 billion liquidation

Author: Gu Yu, ChainCatcher

Born in 2001, this genius teenager entered Columbia University at the age of 15 and graduated at 19 as the top of his class, earning degrees in mathematics, statistics, and economics. Leopold Aschenbrenner was once a core member of OpenAI's super alignment team but was fired due to a security memo. He then founded a hedge fund and grew its management scale to over $20 billion within two years.

In July 2026, this 24-year-old, once hailed as the "AI prophet" by Silicon Valley and Wall Street, marked a significant pause in his myth by publicly announcing a 67% loss in a single month and being forced to liquidate about $16 billion in public market positions.

I. Letter to Investors: "We Disappointed You This Month"

On the last day of July, Aschenbrenner sent a brief yet heavy letter to his fund's LP investors.

"We disappointed you this month," he wrote in the letter.

Behind this letter was the Situational Awareness fund's staggering loss of about 67% in July. According to the investor letter, even accounting for this month's losses, the fund's year-to-date return was still around 80%—but just at the end of May, this figure was approximately 270%. In just two months, glory and collapse were separated by a plunge in the AI sector.

The losses forced Aschenbrenner to make a difficult decision: to sell off the vast majority of his public market positions to Citadel, owned by Ken Griffin, to raise funds to meet margin calls from lenders. Reports indicate that within less than 36 hours, Citadel acquired all of Situational's approximately $16 billion public market positions at a discounted price.

Meanwhile, Situational had reached an agreement late Wednesday to sell its stake in AI giant Anthropic to a consortium of investors led by Greenoaks and Sequoia Capital for $3.5 billion—but by Thursday morning, the fund ultimately withdrew from the deal.

Aschenbrenner had his own interpretation in the letter. He attributed part of the blame to short sellers targeting the fund's positions, claiming they exacerbated the losses, and likened Situational's experience to a "bank run." He also informed investors that the fund had removed all leverage from its portfolio.

He made it clear: "I take full responsibility for these events." The fund was not closed, liquidated, or turned into a pure private equity fund; it would continue to operate in a mixed manner, but public positions would be temporarily fully paid. He promised to strengthen portfolio management, risk teams, and vigilance, turning "expensive scars" into valuable lessons, and expressed optimism about current opportunities—fundamentals were improving rapidly, and prices had significantly dropped.

II. The Collapse: AI Stocks Plunge Combined with Leverage Backfire

The collapse of Situational was not without warning.

Unlike most diversified hedge funds, Aschenbrenner had almost placed all bets on the AI revolution since founding the fund. He had publicly stated multiple times that AI would be the largest investment opportunity globally in the coming years, thus the fund heavily invested in the AI industry chain, including AI computing power, data centers, and power infrastructure, while shorting some traditional software companies he believed would be impacted by AI, hoping to achieve excess returns through a strategy of "going long on AI and shorting traditional software."

During the sustained rise of AI stocks, this strategy performed exceptionally well. By the end of May, Situational's year-to-date return had reached approximately 270%, and its management scale rapidly expanded to over $20 billion, making it one of Wall Street's most watched star funds this year.

However, the problem was that the fund was built on a high leverage foundation. When the AI sector experienced a significant adjustment in July, many heavily held stocks fell simultaneously, while some previously established short positions in software stocks rose against the trend, putting pressure on both sides. As the fund's net value rapidly shrank, lending institutions began to demand margin calls, forcing Situational to sell off some assets to supplement margin, which further depressed stock prices, creating a vicious cycle of "stock price decline—margin call—forced selling—continued decline."

More critically, many of the stocks held by Situational had limited liquidity. When the market realized the fund was facing deleveraging pressure, some institutions began to establish short positions against its heavily held stocks, further amplifying market volatility. Aschenbrenner later likened this event to a "bank run" in the investor letter, believing that when all counterparties attempted to exit first, liquidity would quickly dry up.

Ultimately, the fund did not wait for the market to stabilize. Just a day after Situational completed large-scale deleveraging (July 31), the AI sector experienced a strong rebound, with SanDisk rising 25.99% in a single day and CoreWeave rising 21.51%. This moment became the most dramatic part of the entire event: Aschenbrenner did not misjudge the long-term value of AI; rather, under high leverage, he was prematurely eliminated by short-term market fluctuations.

III. The "Past and Present" in the Crypto World

Few know that this "AI stock god" was once a part of the crypto world.

In early 2022, Aschenbrenner joined the Future Fund team under FTX, founded by Sam Bankman-Fried. FTX was once the second-largest cryptocurrency exchange in the world, with a valuation exceeding $32 billion at one point. At that time, 19-year-old Aschenbrenner was involved in the "effective altruism" (EA) movement, assessing funding projects, hoping to invest hundreds of millions of dollars into areas with the greatest potential to improve human welfare.

However, in November 2022, FTX collapsed due to the misappropriation of customer funds and debt issues and subsequently filed for bankruptcy. Founder SBF was ultimately sentenced to 25 years in prison, and the ethics he touted of "earning to give" were later described by himself as "mostly a front."

Aschenbrenner left the institution just before FTX's collapse. This experience at the center of the world's largest financial scandal made him acutely aware of the complexities and risks of the market.

But the irony of fate is that after leaving FTX, Aschenbrenner joined OpenAI's super alignment team to study AI safety—this job was originally meant to "hit the brakes" on superintelligence. In April 2024, he was fired for writing a security memo warning about OpenAI's inadequate security, under the pretext of "leaking." A month later, the super alignment team also announced its dissolution.

Just two months after being fired, Aschenbrenner published a 165-page article titled "Situational Awareness: The Next Decade." He then turned around and founded the most aggressive AI acceleration fund in the market, betting everything on the AI surge—from someone who was "hitting the brakes" to becoming the one "pressing the accelerator" the hardest.

However, Aschenbrenner has not severed all ties with the crypto world. On the contrary, he has identified the intersection of AI and crypto infrastructure as an important investment direction.

According to SEC filings, the Situational Awareness fund has established a large position in the Bitcoin mining sector. Its holdings include Bitcoin mining companies such as IREN, Core Scientific, Riot Platforms, CleanSpark, and Bitfarms. By the first quarter of 2026, the fund's public holdings had expanded from $5.5 billion to $13.67 billion.

Aschenbrenner's investment logic is not based on the expectation of rising Bitcoin prices but rather on recognizing a deeper structural opportunity: the industrial-grade power infrastructure and grid access rights held by Bitcoin miners.

In the United States, new data centers typically require three to five years to connect to the grid. Bitcoin miners already possess ready-to-use gigawatt-level industrial electricity capacity and long-term energy supply agreements. As AI companies compete for power capacity, the grid access rights held by miners are becoming more valuable than Bitcoin itself.

IV. Leverage, Cycles, and Circle Tightening

The dramatic rise and fall of Situational Awareness reflect Silicon Valley's extreme style of betting on future technologies amid the global AI wave.

Fu Peng, chief economist at Bitfire Group, recently pointed out that the AI industry is currently in a transitional phase from midstream to downstream. The infrastructure layer, represented by Nvidia, has entered a mature stage, exhibiting characteristics of "old tech stocks," while the downstream application layer has yet to see a milestone breakthrough similar to ChatGPT. The next 10 to 18 months will be a transitional period for the industry cycle. The second half of AI's competition will no longer be about burning money to build infrastructure but rather whether AI, as an efficiency tool, can generate stable payments, profits, and diffusion after entering real industries.

Fu Peng also emphasized the risks of leverage: when the fundamentals of the industry are extremely certain and volatility is very low, the excessive accumulation of off-market financial leverage often becomes the invisible killer that triggers a crash. When deleveraging occurs, even if a company's fundamentals are sound, prices can still fluctuate violently.

This is precisely the lesson that Aschenbrenner bought with $20 billion.

Fu Peng also noted that current market funds are following a circle-tightening logic, prioritizing the clearing of peripheral, high-elasticity, and cash flow-deficient speculative assets, continuously concentrating on the most core and highest certainty targets. Pure liquidity assets like Bitcoin often digest valuation pressure before traditional stock markets. Aschenbrenner's forced liquidation may not only signify the fall of a genius but could also become a watershed moment in this round of the AI market.

Aschenbrenner acknowledged in the letter that the fund had removed all leverage. But this decision came too late—once leverage begins to unwind, its destructive power becomes irreversible.

This 25-year-old, emerging from the ruins of FTX, leaving amidst the controversies at OpenAI, created one of Wall Street's most astonishing wealth myths in two years, only to destroy it with his own hands in six days. His story is both a footnote to the fervor of the AI era and a vivid warning against leverage and greed.