Leopold Aschenbrenner wrote his investors a letter on July 24. The market was falling. His biggest positions were bleeding. He told them this was the best buying opportunity since early 2025.

He was asking for more money.

Nobody sent it.

Eleven days later, his fund's entire public stock portfolio was gone — sold in a single block trade to Ken Griffin's Citadel, at prices low enough that the same stocks rallied 20-27% the very next day.

This is how a 24-year-old went from AI's most talked-about investor to the center of one of the wildest blowups Wall Street has seen this year.

The Essay That Made Him A Name

Before the fund, Aschenbrenner worked on OpenAI's superalignment team — the people responsible for keeping powerful AI systems aligned with human intent.

In 2024 he left, and published "Situational Awareness" — a sweeping essay on how close AGI really was, and why almost nobody was taking the timeline seriously. It went viral. It made him a name in AI circles most 23-year-olds never get close to.

Then he did something almost nobody does with a viral essay. He turned it into a fund.

$225 Million → $45 Billion

Situational Awareness LP launched in 2024 on one high-conviction thesis: the AI buildout was real, and the market hadn't priced in how much infrastructure it would take to get there.

Long the picks-and-shovels of AI — semiconductor names like SK Hynix, infrastructure plays like CoreWeave and Nebius. Short the names he thought AI would hollow out — including Adobe.

Sized with leverage reportedly up to 4x. On both legs.

It worked. Spectacularly.

By some accounts the fund started around $225 million. By July 2026 it hit $45 billion — up 439% net of fees. Aschenbrenner was being called one of the best macro traders alive.

Then the same conviction that built the fund started tearing it apart.

The Reversal

Mid-July: his four biggest positions each fell more than 35% in weeks. The Adobe short backfired too — it rallied instead of rolling over.

On a book levered 4x both ways, that doesn't just hurt. It compounds.

July 24: he wrote investors, called the selloff the best entry point since early 2025, asked for fresh capital. The money didn't come.

July 28: Bloomberg reported Citadel Securities — Ken Griffin's market-making arm — was calling for a surprise Fed rate hike. Not priced in. Tech and AI names, already wobbling, sold off harder.

For a fund levered 4x into exactly those names — gasoline on a fire already burning.

The Morning It Ended

July 30. Bank of America, Goldman Sachs, and JPMorgan all issued margin calls the same morning.

No negotiating with three margin calls at once.

Situational Awareness sold its entire public equity portfolio — SK Hynix, CoreWeave, Nebius, all of it — in one block trade, to one buyer: Ken Griffin's Citadel.

Trades this size rarely happen in the open market — dumping billions at once would tank the price further. They get arranged off the tape, negotiated directly between two parties. Which puts enormous leverage in the hands of whoever's writing the check.

The next day, the Fed didn't hike. The same stocks Citadel had just bought rallied 20-27%.

"Being Right And Staying Solvent Are Two Different Skills"

That line has been everywhere in finance circles since the deal broke. It says more than most of the coverage.

Because Aschenbrenner wasn't wrong. The AI buildout he bet on is still happening. His July 24 letter calling the dip a buying opportunity may prove right in hindsight.

Leverage doesn't wait for hindsight. It wants cash today, on the worst possible day — right or wrong.

Here's what's actually driving the conversation right now: Citadel Securities called for a hike that didn't happen, four days before the margin calls hit. The stocks Citadel then bought at distressed prices rallied double digits the moment the hike failed to materialize.

Traders, strategists, and finance accounts across every platform are asking the same question: was this just markets being brutal — or a firm with market-moving influence positioning itself to buy exactly what it knew would be for sale?

No reporting has established coordination. Citadel hasn't been shown to have done anything outside the normal mechanics of market-making and distressed-asset buying. But the optics — forecasting the shock, then buying the wreckage days later — are exactly why more than one strategist has called this "the story of 2026" in the last 48 hours.

What's Left

The fund isn't gone. It still holds private stakes — reportedly including Anthropic. Aschenbrenner, barely 24, is still sitting on returns most of Wall Street would envy on a longer horizon.

The public book — the one built on leverage, conviction, and a letter that came six days too early — is what didn't survive.

The lesson everyone in this industry is drawing from it, whether they say it out loud or not: being early and being right can look identical to being wrong, right up until your broker calls asking for cash you don't have.

Based on reporting from CNBC, The Wall Street Journal, The New York Times, Reuters, Bloomberg, the Financial Times, and MarketWatch, alongside public commentary circulating across finance and social media. DeepTape will continue following this story as it develops.