He signed a 135-page essay and then wrote checks big enough to make people stare. You watched as those bets, once a thesis on the future of AI, turned into urgent phone calls and selling orders. I remember the afternoon a desk lit up with margin alerts—there’s a particular quiet that follows that sound.
I’ve followed founders who become fund managers long enough to know the moves that look brilliant on paper can feel reckless in a market squeeze. You can admire the audacity and still ask how someone running a firm named Situational Awareness ended up in a firefight on two fronts.
SK Hynix and Adobe printed different headlines on the same day.
The chip supplier sank while Adobe climbed more than 30% in a month—two tradeable stories that collided. Leopold Aschenbrenner’s Situational Awareness held big long positions in AI infrastructure names (SK Hynix among them) while simultaneously placing shorts against software incumbents like Adobe, betting AI would compress their markets. When the infrastructure names slumped and Adobe rallied, the fund faced the kind of cash strain that forces rapid liquidation.
What happened to Leopold Aschenbrenner’s fund?
CNBC reported the fund had to sell the majority of its public positions after suffering steep losses. Situational Awareness reportedly managed about $45 billion (≈€41 billion) at the start of July—after a year that included a headline 2,000% return for 2025, according to coverage—and then shrank fast as positions were offloaded, largely to Citadel, the firm run by Ken Griffin.
Traders on the floor saw buyers circle before the exits opened.
When speed matters you sell to the quickest bidder. The Financial Times says Situational Awareness had other potential buyers but chose Citadel to limit further slippage; Citadel moved quickly and some of the sold assets staged rallies afterward. That rapid trade execution felt less like careful risk management and more like putting out fires with gasoline.
Why did Situational Awareness sell its positions?
Margin pressure and asymmetric losses. Long exposure to volatile memory and chip stocks collided with short exposure to entrenched software firms that held up—or rallied—faster than expected. Short positions require counterparties and margin; when they turn against you, options narrow and cash calls arrive.
In dark rooms, private stakes didn’t get a headline, but they mattered.
Public liquidation doesn’t equal personal ruin. Reports in Business Insider say Aschenbrenner kept sizable private holdings, notably in Anthropic, which is eyeing a public listing later this year. That private exposure, plus his personal ties to Anthropic’s leadership via his engagement to Avital Balwit, means he didn’t exit the game—he re-leveraged the table behind closed doors.
Is Aschenbrenner still invested in AI companies?
Yes. While nearly all public positions were sold, large private positions reportedly remain. If Anthropic goes public, those stakes could revalue dramatically—so the headline losses may prove only a pause in a larger wealth story.
On social feeds and regulator pages, reputations are rewriting themselves.
People are talking about Citadel as more than a buyer: Reuters reminded readers of past flashpoints—Robinhood and GameStop—and the optics of a large market player scooping up distressed books has political and PR consequences. You’ll see chatter on X and in newsroom Slack channels debating whether this was a rescue, a takeover, or a market-clearing event.
I’ve watched market narratives bend around a single personality before. Leopold built a thesis and bet the firm on it; that’s bold, and boldness can look like genius or like a house of cards when the wind changes.
You want lessons? Keep your perspective calibrated: big public losses can mask private optionality, and cultural signals—an engagement, an IPO rumor, a name in the press—often matter as much as price charts. Markets punish binary bets that forget reality’s messy middle.
So what do you think: was this a brutal comeuppance for overconfidence, or simply a tactical reset by someone who never intended to play only in public markets?