📰 Key Takeaways

The Situational Awareness hedge fund, founded by former OpenAI researcher Leopold Aschenbrenner, nearly collapsed in July after an AI stock selloff triggered margin calls. It has since put another $400 million into an unnamed private company, with the deal closing this past Tuesday — on top of a $100 million investment in the same company back in July. Bloomberg reported the news on Thursday, citing people familiar with the matter.

July’s AI stock selloff triggered margin calls from Wall Street lenders, causing the fund’s assets to shrink by roughly 78%. According to the Financial Times, the fund sought fresh capital at the time and considered selling off its private company stakes to stay afloat. It ultimately sold most of its public equity portfolio to Ken Griffin’s Citadel, using the proceeds to pay down lender debt while holding on to its private positions.

The fund has historically made big bets on power and data-center plays to back AI’s growth, including Bitcoin mining companies pivoting into AI compute. Per an SEC filing dated May 18 covering positions as of March 31, the fund held roughly $1.11 billion across seven Bitcoin mining stocks, including IREN, Core Scientific, Riot Platforms, and CleanSpark.

Cointelegraph has reached out to Situational Awareness for comment and has not yet received a response.

💬 JudyAI Lab Take

What’s worth paying attention to here isn’t the dollar figure — it’s what this reveals about a risk layer that often gets overlooked: when “AI conviction” itself becomes a highly leveraged position, a shift in market sentiment hits the infrastructure funding that underpins AI development directly.

This fund had piled into power, data centers, and Bitcoin mining stocks pivoting into AI compute — a classic “bet on the long-term AI narrative” strategy. But July’s AI stock selloff triggered margin calls and wiped out nearly 80% of assets, forcing the fund to sell public positions just to survive — and it still chose to double down with $400 million in a private company afterward. That tells you something about a design tradeoff: public market positions can’t absorb short-term volatility under leverage, while private investments are relatively more “shock-resistant” precisely because they lack real-time pricing. For AI builders, the lesson is that when you’re evaluating any AI-related investment or partnership, it’s not enough to ask whether the narrative holds — you also need to ask whether the capital structure can survive the turbulence in between.

Next time you’re sizing up a partner or investment target in AI, it’s worth asking one more question: can their leverage structure withstand a short-term market headwind?

📅 Source Info

🔗 Further Reading