AI hedge fund run by former OpenAI researcher Leopold Aschenbrenner plunges from $45 billion to $10 billion

Two years ago, Leopold Aschenbrenner argued he was one of few people in the world who saw the future clearly.

In a sprawling, 165-page essay that became required reading in Silicon Valley, the former OpenAI researcher positioned himself as a kind of prophet for the coming age of artificial super intelligence.

But this week, the limits of Aschenbrenner's vision were on display when the AI-themed hedge fund he runs — named Situational Awareness, also the title of his viral June 2024 manifesto — ran into the harsh reality of tumbling semiconductor stocks and Wall Street margin calls.

At its peak earlier this month, his fund sat atop $45 billion in assets. By Thursday, however, after being forced to offload all of his leveraged stock bets — including hard-hit names like SK Hynix and CoreWeave — to Ken Griffin's Citadel at a discount, the fund's holdings plunged to around $10 billion, according to people with knowledge of the situation.

The story of Aschenbrenner's meteoric rise and sudden fall has captivated both Wall Street and tech circles, making him the most high-profile casualty yet of the volatility accompanying the AI boom.

A polarizing figure, his online followers saw Aschenbrenner — a Columbia University valedictorian at age 19 — as a genius of the next big thing and followed his fund's quarterly filings for clues on hot AI stocks.

Before this month's decline, Situational Awareness racked up gains of more than 1,000% since inception, The Wall Street Journal reported last month. The Journal noted Aschenbrenner was just 24 years old.

Meanwhile, critics pointed out that Aschenbrenner had no experience running money prior to launching his fund in July 2024, calling him more lucky than smart. Some noted that his early work experience was at the doomed crypto firm FTX, where he helped now-disgraced founder Sam Bankman-Fried run a charity out of a Bahamas penthouse.

Others on Wall Street, including former traders at global investment banks, noted that in light of reports Situational Awareness used as much as 400% leverage, the collapse wasn't shocking.

"A lot of people saw this blow-up as a matter of not if, but when," said Jerry Diao, who runs a Wall Street coaching firm. "Maybe his views on AI are correct in the long run, but in the public markets, you have to be prepared for the short-term."

The hedge fund didn't immediately respond to a request for comment from CNBC.

Earlier this week, before the sale to Citadel, about two-thirds of Situational Awareness holdings were in long and short positions in public equities, according to one source. The rest were stakes in private companies, dominated by a multibillion-dollar Anthropic investment, the person said.

CNBC's sources spoke on the condition of anonymity to discuss nonpublic details.

The near-collapse of Situational Awareness coincides with the hedge fund manager's wedding, set for this weekend, sources told CNBC's David Faber. Aschenbrenner is engaged to Avital Balwit, chief of staff for Anthropic CEO Dario Amodei, according to a Fortune profile.

'Weirdness' and 'disagreeableness'

Born in Germany to physician parents before moving to the U.S., Aschenbrenner showed an early aptitude for math and computer science, according to profiles and podcast interviews.

He skipped several grades in the German school system, graduating high school at age 15, and as a teen at Columbia University he garnered attention for an academic paper titled, "Existential Risk and Growth."

A Columbia classmate, Sofia Montrone, said that she hadn't heard of Aschenbrenner before meeting him over Zoom shortly before their 2021 graduation.

"It was not like he was some prince, emerging out of the school," Montrone told CNBC. "He was just some guy."

In the interaction, Montrone, who was salutatorian, said she found her classmate "child-like" and socially awkward.

Aschenbrenner has since said that his personality — what he called his own intellectual "weirdness" and "disagreeableness" — was punished in German culture. He came to see it as the source of his edge.

While at Columbia, he co-founded the school's chapter of Effective Altruism, a philosophy popular in some tech circles that advocates for founders to make the most money possible in order to help humanity.

That network became his career pipeline, eventually leading him to work with another effective altruism proponent — Bankman-Fried — after his graduation in 2021. He worked for a stint at the Future Fund, the philanthropic arm of FTX, before the crypto firm's collapse.

In 2023, Aschenbrenner landed on OpenAI's Superalignment team, working under Ilya Sutskever on the problem of keeping AI aligned with human interests. After a hacker breached OpenAI's internal systems, he wrote a memo to the board warning that the company's security wasn't strong enough to stop foreign espionage, naming China specifically.

In 2024, the company fired Aschenbrenner after accusing him of improperly sharing confidential information, a characterization he has disputed, saying he was raising concerns about the company's security practices.

"I liked Leopold while at OpenAI," Scott Aaronson, a computer scientist now at the University of Texas at Austin who previously worked on AI safety at OpenAI, told CNBC this week in an email.

"I was sorry when he got pushed out because of sharing information in a way leadership didn't approve of," he said. It "sounded like he was trying to do the right thing and they overreacted."

An OpenAI spokesman declined to comment and referred to statements the company made at the time that the firm disagreed with many of Aschenbrenner's claims.

Representatives for Columbia University and its Effective Altruism chapter didn't respond to requests for comment.

Stripe, Github investors

Weeks after his departure from OpenAI, Aschenbrenner turned his brief experience at the leading AI firm into a sweeping vision of where artificial intelligence, and the world, was headed.

His June 2024 essay argued that artificial general intelligence could arrive within years and that governments were badly underestimating the pace of progress. Admirers saw it as evidence that Aschenbrenner was a prodigy with valuable insight into AI's trajectory, while critics said it overstated both the technology's near-term capabilities and his own certainty about the future.

By July of that year, Aschenbrenner parlayed his rising fame into seed capital for his hedge fund, raising a reported $225 million from Stripe co-founders Patrick and John Collison, former GitHub CEO Nat Friedman and investor Daniel Gross.

That would mark the start of a two-year run unlike any in recent Wall Street history.

"Before long, the world will wake up," Aschenbrenner wrote at the time, adding that only a few hundred people in the AI community knew what was coming.

"If they are seeing the future even close to correctly," he wrote, "we are in for a wild ride."