The youngest hedge fund founder to blow up $20 billion in public equities still owns the one asset Wall Street can't price.
The Summary
- Leopold Aschenbrenner's AI-focused hedge fund Situational Awareness was forced to sell most of its public equity portfolio to Citadel after leveraged bets plummeted, but the firm still holds its Anthropic shares
- The mid-20s former OpenAI researcher went from Columbia valedictorian at 19 to Silicon Valley darling to the center of a high-profile fund meltdown in less than five years
- The public portfolio collapsed, but the private Anthropic stake, illiquid and unpriced, represents either total loss or the comeback trade of the decade
The Signal
Leopold Aschenbrenner graduated from Columbia at 19 as valedictorian, joined OpenAI's superalignment team, got publicly fired, then launched a hedge fund betting billions on the same AI trajectory that made him unemployable at his last job. The fund, named Situational Awareness after the very capability he was supposed to help build into superintelligent systems, hit $20 billion in assets under management before it hit a margin call.
The firm sold most of its public holdings to Citadel after leveraged positions in public AI equities went underwater. This is the part making headlines. But the fund still holds its Anthropic shares, the private equity stake that can't be liquidated in a fire sale because there's no public market for it.
"The former OpenAI researcher still has cards to play."
Here's what that means. The public portfolio was the leveraged bet. Buy Nvidia, Microsoft, the whole AI trade everyone else saw coming, and juice returns with borrowed money. When those bets reversed, the fund had to unwind or face total wipeout. Citadel, always ready to buy distressed assets from people who thought they were smarter than the market, picked up the pieces.
But Anthropic, OpenAI's primary competitor and the company building Claude, is still private. Aschenbrenner worked at OpenAI, knows the model training game from the inside, and presumably saw something in Anthropic worth betting on when the shares were available only to insiders and early-stage funds. That position is illiquid, meaning he can't sell it even if he wanted to. It's also unpriced, meaning no one knows what it's worth until Anthropic either goes public, gets acquired, or goes to zero.
Key questions the meltdown raises:
- How much leverage did a mid-20s founder get on a fund predicated entirely on AI hype?
- Who gave him that leverage, and what did they think they were buying?
- What's the Anthropic stake actually worth if the thesis was right but the execution was early?
Aschenbrenner co-founded an effective altruism group at Columbia, the same philosophical movement Sam Bankman-Fried used to justify his own leveraged bets on the future. That's not guilt by association, but it is a pattern. Effective altruism attracts people who believe they can see the future more clearly than the market, and that conviction makes them double down when they should pull back.
The irony is thick. Situational awareness, in AI terms, means a model knows what's happening around it and can adjust. Aschenbrenner's fund, named for that capability, apparently lacked it when the margin calls started.
The Implication
If Anthropic's next funding round or IPO values the company higher than the price Aschenbrenner paid, he walks away from this meltdown as the guy who saw the future but mistimed the trade. If Anthropic stalls or the AI model arms race consolidates around OpenAI and a few others, the private stake becomes a tax write-off and a cautionary tale. Either way, this is the shape of capital formation in the agent economy: young, leveraged, convinced of inevitability, and one margin call away from irrelevance.
Watch what happens to the Anthropic shares. That's the real tell on whether the thesis was wrong or just early.