Leopold Aschenbrenner, the former OpenAI researcher whose 2024 essay "Situational Awareness" argued that artificial general intelligence could arrive around 2027, turned that thesis into a hedge fund of the same name. According to reporting from CNBC, Bloomberg and the Financial Times, it had been one of the year's standout performers — the FT, citing a July 24 investor letter, reported a net return of 439% in the first half of 2026 and more than 1,000% since inception.
That run reversed sharply in the final week of July. Multiple outlets reported that a selloff in semiconductor and AI-linked stocks, amplified by leverage the reports put as high as roughly 400%, triggered margin calls that forced the fund to unwind. CNBC reported on July 30 that Situational Awareness sold its leveraged book of public positions — names cited include SK Hynix and CoreWeave — to Ken Griffin's Citadel, reportedly at a discount. Reports described assets falling from around $45 billion to roughly $10 billion; the figures vary between outlets and the fund has not publicly confirmed them.
The episode is being read less as a verdict on the AGI thesis itself than on the machinery around it. A concentrated, highly leveraged position in AI infrastructure behaves like the AI trade — only faster, and in both directions. When the underlying stocks fell, leverage compressed a drawdown into a forced sale measured in days rather than quarters, and a larger rival was positioned to buy the book. For a fund premised on the long-run trajectory of intelligence, the near-term reminder is an old one: leverage sets the timetable, and the timetable is not always yours.
This is market analysis, not a recommendation to buy or sell any security.