Citadel Securities Acquires $10B AI Infrastructure Portfolio from Situational Awareness LP

When a hedge fund goes from zero to $45 billion in assets under management in under two years, the story has exactly two possible endings. On July 30, 2026, Situational Awareness LP found out which one it got.

Citadel Securities paid $10 billion to acquire the public equity holdings of Situational Awareness LP, the AI-focused hedge fund founded by former OpenAI researcher Leopold Aschenbrenner. The deal hands Ken Griffin’s firm a portfolio built almost entirely around AI infrastructure bets, at a moment when those bets have gone badly wrong.

Situational Awareness launched in 2024 and was backed by high-profile tech investors including the Collison brothers and Nat Friedman. The fund assembled a concentrated portfolio of AI infrastructure plays: data centers, power supply companies, semiconductors, and Bitcoin miners pivoting toward high-performance computing.

At its peak, the fund’s assets under management reached nearly $45 billion. Shortly before the sale, that figure had compressed to over $20 billion. The publicly traded portion of the portfolio, valued somewhere between $10 billion and $16 billion at the time of the transaction, sat at the center of the problem. Situational Awareness had financed a significant chunk of its positions with leverage. When AI-related equities turned south in mid-2026, margin calls started arriving, and the fund needed liquidity fast.

Among the holdings were Bitcoin miners Core Scientific, ticker CORZ, and Iris Energy, ticker IREN, two companies that had repositioned themselves as high-performance computing infrastructure providers for AI workloads. The fund held roughly $1 billion in Bitcoin mining stocks within a broader $5.5 billion equity portfolio prior to the full sale. Citadel’s acquisition reportedly included the debt-financed portion of Situational Awareness’s leveraged positions, not just the underlying equity.

For Situational Awareness, the transaction buys survival, or at least an orderly wind-down. A fund facing simultaneous margin calls and falling NAV had limited options: sell assets at distressed prices, negotiate with lenders, or find a single counterparty willing to take the whole position off the table. Citadel became that counterparty.

Aschenbrenner had built his public thesis around the idea that AI infrastructure was the most important capital deployment opportunity in a generation. His 2024 essay series on artificial general intelligence timelines attracted significant attention and helped seed the fund’s early fundraising.

For Bitcoin miners specifically, the repricing is notable. Core Scientific and Iris Energy had spent the prior two years convincing the market to value them as AI infrastructure companies rather than crypto miners. A forced seller of that magnitude hitting their stock creates a real test of whether that re-rating holds.