Leopold Aschenbrenner's Situational Awareness fund seeks fresh capital after AI selloff hammers its bets

Leopold Aschenbrenner's Situational Awareness LP hedge fund, which grew from $225 million to $20 billion on a pure AGI bet, is seeking fresh capital after the July 2026 semiconductor rout hammered its concentrated positions in CoreWeave, Bloom Energy, and crypto miners, according to Bloomberg and the Financial Times. The fund's Q1 2026 13F filing showed $8.47 billion in put exposure against semiconductor companies, but both its long and short books suffered as the AI stock selloff accelerated following Meta Platforms' July 1 announcement of its Meta Compute cloud unit.

The hedge fund that turned $225 million into $20 billion on a pure AGI bet is now in talks with investors for a lifeline after the July 2026 semiconductor rout punished its most concentrated positions. Leopold Aschenbrenner built his reputation on a single, blunt argument: AGI is coming faster than anyone admits, and the infrastructure around it is the trade of the decade. That thesis made him famous, then very rich. Now it's causing serious pain. Situational Awareness LP, the hedge fund Aschenbrenner launched in September 2024 at age 22, is seeking new capital after losses tied to the broader AI stock selloff that has rattled markets through July, according to reporting from the Financial Times and confirmed by Bloomberg on July 30. The fund had reached roughly $20 billion in assets under management, an almost incomprehensible trajectory from its $225 million seed round backed by Patrick and John Collison, Daniel Gross, and Nat Friedman. Jane Street had joined as an institutional investor earlier this year, and the Wall Street Journal reported the fund up 270% after fees in 2026 through May. It was, by almost any measure, the hottest fund in the world. Then the AI trade cracked. The fund's disclosed portfolio explains the damage clearly. Its Q1 2026 13F filing showed $8.47 billion in put exposure against semiconductor companies, a short book betting against the chipmakers at the core of the AI buildout. Alongside that, the fund held long positions in crypto miners as infrastructure proxies, Bloom Energy as a power play, and CoreWeave as the GPU rental market's pure-play. Every one of those longs got hit. The July 2026 semiconductor selloff was partly triggered on July 1 when reports surfaced that Meta Platforms was planning to launch Meta Compute, a cloud unit that would sell surplus AI training capacity to enterprise customers. That single announcement rewrote the supply-demand story for GPU infrastructure overnight. CoreWeave, already down roughly 43% from its peak, fell another 35% in mid-July when Meta's commercial cloud service became official. Bloom Energy dropped 15% in a single session around the same period and shed 40% over the course of one month, despite a $14 billion contracted backlog and Q1 revenue up 130% year over year. Bitcoin mining stocks sank approximately 20%, as investors who had been treating miners as AI infrastructure proxies suddenly re-evaluated whether that framing still held. Aschenbrenner's short book, those $8.47 billion in semiconductor puts, was supposed to hedge exactly this kind of move. The problem is the timing and the magnitude. When a fund is long infrastructure names that behave like semiconductor proxies and short the semiconductors themselves, a disorderly rotation in one direction doesn't always produce the clean offset you'd model. Both sides can hurt at once if the long book bleeds faster than the short book gains. According to the Bloomberg report, the fund is in discussions with existing investors and lenders about raising capital. Some investors were offered the opportunity to buy portfolio assets directly, a sign that the fund may need liquidity at the position level, not just at the fund level. That distinction matters. Selling portfolio assets to investors rather than simply drawing on credit lines suggests the fund needs to reduce exposure, not just bridge a gap. Is this strategy failure or a rough patch? Frankly, that question is still open. The fund's broader thesis hasn't been proven wrong. Aschenbrenner's 2024 essay series, which predicted AGI around 2027 and a historic surge in compute and energy infrastructure to support it, wasn't a market timing call. It was a decade-horizon structural argument. The infrastructure positions the fund holds - CoreWeave, the miners, Bloom Energy - remain directly tied to that thesis even after the selloff. But the scale of the drawdown, and the speed at which the fund has moved to seek outside capital, tells a different story. A fund up more than 1,000% after fees since launch - less than two years ago - doesn't go looking for emergency capital without reason. The original backers, the Collisons, Gross, Friedman, and later Jane Street, didn't sign up for this. That's not the script anyone wrote in September 2024. What Aschenbrenner's situation illustrates is a problem particular to highly concentrated, thesis-driven funds. When your entire book is an expression of one argument, a sufficiently sharp move against the infrastructure complex doesn't just cost you money. It forces the conversation about whether the argument itself is broken. The fund's public response has been to urge investors to treat the current moment as a buying opportunity, which is exactly what you'd expect from a manager who believes the pullback is temporary noise against a structural signal. Whether existing investors agree is now the real question, and the fact that asset sales are on the table suggests at least some of them don't. The fund's Anthropic stake, reportedly about one-fifth of total assets, is not a publicly traded position and offers a degree of insulation from the daily mark-to-market chaos. But private holdings don't solve a liquidity crunch when the rest of the portfolio is moving against you. That much is simple. Situational Awareness turned a thesis into one of the most dramatic fund launches in recent memory. Whether it survives the first serious test of that thesis intact is now the story. Also read: Tether Signs Deal With Nairobi Securities Exchange to Tokenize African Stocks https://startupfortune.com/tether-signs-deal-with-nairobi-securities-exchange-to-tokenize-african-stocks/ • Zhongji Innolight's record Hong Kong IPO stumbles out of the gate as geopolitical risk reprices AI hardware https://startupfortune.com/zhongji-innolights-record-hong-kong-ipo-stumbles-out-of-the-gate-as-geopolitical-risk-reprices-ai-hardware/ • Samsung's chip division posted a 250-fold profit surge as AI memory shortages rewrite the rules of the semiconductor market https://startupfortune.com/samsungs-chip-division-posted-a-250-fold-profit-surge-as-ai-memory-shortages-rewrite-the-rules-of-the-semiconductor-market/