A Fund Named Situational Awareness Could Not See Inside Its Own Book

Leopold Aschenbrenner’s hedge fund returned 439% in the first half of this year and grew to roughly $45 billion by early July. Today it sold its entire public equity book to Citadel after AI infrastructure longs and software shorts moved against it at the same time, with prime brokers at Bank of America, Goldman Sachs and JPMorgan working to meet the margin calls. Assets are reportedly down to around $10 billion. What is left is mostly a private stake in Anthropic.

I want to be careful here, because this is not a story about one young manager being wrong. He was early on compute demand and he made more money being right than almost anyone alive. What broke him was not the thesis. It was the leverage underneath it, and the fact that both sides of his book were expressions of the same single view.

That is the part worth sitting with. A 439% return is not evidence of insight, it is evidence of concentration, and concentration looks like genius right up until the tape turns.

Watch what happened next, because it tells you more than the loss does. The AI infrastructure names he was long rebounded hard the same day, several of them twenty percent or more. That is not a sector being repriced on fundamentals. That is forced selling clearing. It means the price action we have been reading as a verdict on AI demand has partly been a function of who was levered and who got a call.

In my work the failure mode is almost always the same. Someone maps the risk they own, stops at the boundary of what they control, and then gets taken down by a dependency that never made it onto the register. This is that same failure at market scale.

Here is what worries me for the sectors around this. The commitments being made right now are long dated and physical. Gas generation, transmission, land, multi year leases, credit guarantees standing in for tenants who cannot borrow on their own name. All of it gets built on the assumption that the demand signal is real. The equity capital validating that signal turns out to be fast money that can disappear inside a week.

I do not know how thin the bubble is. Nobody does, and anyone quoting you a number is selling something. But I know the difference between a demand signal and a leveraged bet on a demand signal, and I do not believe the people underwriting the physical buildout are reliably telling those two apart.

If a fund named Situational Awareness could not see the risk sitting inside its own book, what makes us confident the utilities, lenders and boards financing this buildout can see theirs?