1 hour ago
9
Leopold Aschenbrenner just wrote a $400 million check to a private company backed by Sequoia Capital. Sequoia partner Alfred Lin confirmed the investment on August 6. The identity of the recipient company remains undisclosed, though the deal lands squarely in the AI startup ecosystem.
From $45 billion to $10 billion in a month
Aschenbrenner’s hedge fund, Situational Awareness LP, just went through a significant decline. The fund’s assets under management peaked at roughly $45 billion in July 2026. By late in the month, that figure had cratered to approximately $10 billion. The fund lost about 78% of its asset base in a matter of weeks.
The culprit was a cocktail of heavy leverage, margin calls, and a sell-off in AI infrastructure stocks. The fund was forced to liquidate a majority of its public equity holdings just to stop the bleeding.
The $400 million investment represents a 4% concentration bet relative to the fund’s current size of $10 billion.
The Aschenbrenner playbook
Leopold Aschenbrenner is a former OpenAI Superalignment researcher who published a sprawling essay titled “Situational Awareness” in 2024. The piece predicted rapid advancements toward artificial general intelligence. The fund reportedly delivered 200% net returns for some investors in 2025, with a 47% gain in just the first half of that year.
AI drug discovery keeps attracting capital
While the specific company receiving Aschenbrenner’s $400 million hasn’t been named, AI drug-discovery startups have been pulling in substantial funding rounds throughout 2026. Chai Discovery closed a $400 million round on July 14, 2026, with participation from Sequoia Capital, Index Ventures, and Kleiner Perkins. Aschenbrenner has been linked to potential involvement in that round, though nothing has been formally confirmed connecting the two.
Traditional pharmaceutical development takes over a decade and costs billions per approved drug. AI promises to compress both the timeline and the cost by simulating molecular interactions, predicting drug candidates, and identifying targets that human researchers might miss.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·