A week ago, Leopold Aschenbrenner’s hedge fund, Situational Awareness, was staring down collapse. This week, Silicon Valley is trying to hand him more money.
Investors have been asking to put fresh capital into Situational Awareness, and Aschenbrenner has been turning them down the fund isn’t taking new money right now. It’s an odd position to be in, considering what just happened.
The run-up was staggering. Aschenbrenner told investors in a July letter that Situational Awareness had returned 439% in the first half of 2026 alone.
Money followed the returns. Jane Street, a firm that almost never puts money into outside managers, decided to become one of Aschenbrenner’s investors during that stretch. For someone who left OpenAI with an essay and no trading history, that’s about as strong a vote of confidence as Wall Street hands out.
Then the bets that built that return started working against him. Situational Awareness had leaned hard into AI infrastructure names chips, data centers, power and when those stocks turned, the leverage behind them turned a rough patch into a crisis. Lenders started calling in collateral. Aschenbrenner had to sell almost everything the fund held in public markets, and Ken Griffin’s Citadel was there to buy it, at prices that only make sense for a seller with no room left to negotiate. Situational Awareness went from a peak of $45 billion to roughly $10 billion in the span of days.
What happened next is the part most people are skipping past. Rather than pull back and lick his wounds, Aschenbrenner put $400 million into a private company just days later on top of $100 million Situational Awareness had already committed to the same company the month before. Behind the scenes, the fund’s private bets stretch further than most people realize, touching names like Anthropic and a handful of other AI infrastructure plays that never showed up in the public unwind.
It leaves two very different reactions sitting side by side. On one side, banks are quietly re-examining how much leverage they let AI-focused funds carry, treating this as the warning they’d been waiting for. On the other, the people who believed in Aschenbrenner’s original AI thesis don’t seem to think the thesis broke they think the leverage did, and they’re lining up to bet on Situational Awareness again anyway.
Whether that confidence looks smart or reckless probably depends on what happens the next time the trade turns.
