Just weeks ago, Situational Awareness LP was regarded as a brain trust for the AI boom. Now it is the subject of a catastrophic post-mortem. The hedge fund, founded by 24-year-old Leopold Aschenbrenner, has reportedly sold most or all of its public stock portfolio to Ken Griffin’s Citadel after a punishing decline in AI stocks. According to reports, the fund was worth approximately $45 billion at the start of July. After the asset sale, it is now worth roughly $10 billion. If those figures hold, the fund has burned through about $35 billion in a matter of weeks.

That would make the situation one of the largest trading losses ever recorded. The previous benchmark, Archegos Capital Management, lost $8 billion in ten days in 2021. Situational Awareness appears to have lost multiple times that amount. Every detail of the collapse seems to make the story more pointed. The fund had a staff of eight, and only four of them were investment professionals. Its largest holdings at the end of the first quarter included Nebius Group, Sandisk, Micron, and CoreWeave. All four of those stocks are down more than 35 percent this month.

Key facts at a glance

• Fund founder: Leopold Aschenbrenner, a 24-year-old former OpenAI employee.
• Fund size: About $45 billion at the start of July, now about $10 billion.
• Portfolio sale: Sold most or all of its public stock portfolio to Citadel, Ken Griffin’s hedge fund.
• Holdings: Concentrated in AI-related names including Nebius, Sandisk, Micron, and CoreWeave.
• Backers: Stripe co-founders Patrick and John Collison, Meta AI leaders Daniel Gross and Nat Friedman, and Jane Street.
• Remaining assets: Private holdings, including a reported $5 billion stake in Anthropic.

An AI bet built on a grand theory

Situational Awareness was named after a series of essays Aschenbrenner wrote about machine intelligence. The essays argued that artificial general intelligence was not only possible, but imminent. Aschenbrenner predicted that AI systems would outpace many college graduates by 2025 or 2026, and that superintelligence would arrive by the end of the decade. He described a coming national security project, a race with China, and the possibility of all-out war if the West fell behind. The grandiosity of the claims helped attract attention, but it also turned the hedge fund into a kind of ideological bet.

The hedge fund’s strategy was simple: pour as much money as possible into AI stocks and wait for the future to arrive. Aschenbrenner described the firm as a brain trust on AI. He believed that the fund would have more situational awareness than traditional money managers in New York. “We’re definitely going to do great on investing,” he said in a podcast interview, while also claiming that the firm would be a public voice of reason and an advisor to those making sense of the AI transition.

That confidence appears to have been catastrophic. Leverage was central to the strategy. If an investor borrows money to increase exposure, gains are magnified. Losses are magnified too. At one point, the fund reportedly claimed to be up 439 percent. But when AI stocks wobbled, the borrowed money became a problem. Lenders began issuing margin calls, demanding more collateral. The fund first tried to raise additional capital from investors and lenders, and even offered investors a chance to buy parts of its portfolio. That was not enough. The sale to Citadel followed.

A founder with an unusual resume

Aschenbrenner’s path to hedge fund manager was anything but conventional. At 17, he was called an economics prodigy by Tyler Cowen, a libertarian economist well known in Silicon Valley circles. Cowen’s Emergent Ventures gave him a grant. Aschenbrenner later published essays in Works in Progress, a publication funded by Stripe. While at Columbia University, he co-founded the school’s Effective Altruism chapter. He graduated as valedictorian in 2021 at age 19.

He then joined the FTX Future Fund, the philanthropic arm of FTX, the cryptocurrency exchange founded by Sam Bankman-Fried. FTX later collapsed after a massive fraud was revealed. Among Aschenbrenner’s coworkers at the Future Fund were William MacAskill, a key figure in Effective Altruism, and Avital Balwit, who would later become chief of staff at Anthropic. After FTX, Aschenbrenner joined OpenAI’s superalignment team, which focused on making advanced AI systems behave safely.

His tenure at OpenAI was short. Reports describe him as arrogant, abrasive, and politically clumsy. In one anecdote, he allegedly told then Scale AI CEO Alexandr Wang how many GPUs OpenAI had during a holiday party, in front of a group. Both men denied the exchange took place. Aschenbrenner was eventually fired from OpenAI for leaking internal information in an incident that was unrelated to the GPU conversation. Two months later, he published Situational Awareness, the essay series that became the theoretical foundation for his fund.

There was a notable gap in his resume: no formal money management experience. He had spent a few months at FTX and about a year at OpenAI. Yet investors were apparently willing to overlook that. The fund’s backers included Stripe co-founders Patrick and John Collison, as well as Daniel Gross and Nat Friedman, two leaders in Meta’s AI efforts. The fund’s director of research was Carl Shulman, who had worked at Peter Thiel’s Clarium Capital. Jane Street, the investing firm known for hiring gifted quantitative minds, also bought in. That investment was seen as particularly significant because Jane Street rarely allocates capital to outside money managers.

Why did serious investors buy in?

The obvious question is why apparently sophisticated investors trusted a 24-year-old with no track record. The answer may have more to do with social influence than with financial analysis. Aschenbrenner had carefully cultivated relationships in the Effective Altruism and AI safety communities. He had the endorsement of respected economists and the network of FTX’s academic beneficiaries. The essays were not rigorous investment research, but they were offered as a sweeping vision that flattered certain readers. Axiomatic acceptance of that vision may have felt like being on the inside of a secret future.

Social proof is a dangerous substitute for due diligence. The collapse of Theranos and the Madoff scandal both demonstrated how easily reputation can be confused with competence. In Silicon Valley, the pattern is especially common. A young founder with a confident narrative can attract enormous sums of money from people who trust their own instincts more than a balance sheet. Situational Awareness appears to be another example of that dynamic.

The essays themselves were not subtle. Aschenbrenner wrote that one could “see the future first in San Francisco.” He claimed that AI would create national security forces not seen in half a century. He compared an impending race with China to a contemporary version of the Manhattan Project. For a certain kind of Silicon Valley reader, these statements were thrilling. They made a speculative bet feel like a civic duty. The hedge fund was therefore not just a way to make money; it was a way to align capital with destiny.

That framing may explain why the fund was able to raise so much, so quickly. It may also explain why the downside was not carefully considered. Aschenbrenner acknowledged in a podcast that “obviously, not blowing up is task number one and two.” He said timing was critical, and that the sequence of bets on the way to AGI was underrated. “You’ve got to be really, really careful about your overall risk positioning,” he added, “If you expect crazy events to play out, there’s going to be crazy things you didn’t foresee.” One of those unforeseen events may simply be that artificial general intelligence did not arrive on schedule, and AI stocks went from expensive to overvalued.

The mechanics of the crash

Hedge funds routinely borrow money to amplify their bets, a practice known as leverage. If a manager is confident in a thesis, borrowing can turn a modest gain into a spectacular result. But leverage also turns a modest decline into a forced sale. That is what happened to Situational Awareness. As AI stocks fell, the fund’s lenders demanded additional money to secure their loans. The fund no longer had enough capital to meet those demands without selling assets. Because public equities are the most liquid part of a portfolio, they were the first to go.

The scale of the public portfolio was significant. Reports indicated that Situational Awareness had roughly $16 billion in public equity before the sale. Citadel, Ken Griffin’s giant hedge fund, acquired most or all of it. The terms of the sale were not disclosed. The remaining private holdings include a reported $5 billion stake in Anthropic, one of the leading AI labs. Negotiations to sell that stake were reportedly underway, but it was unclear whether a deal had been completed.

The timing could not have been worse. A few days earlier, market observers had noted a rising sense of anxiety about AI stocks. The sector had been the main driver of stock market gains. When the mood shifted, the stocks fell quickly and steeply. Situational Awareness was not diversified. It was a concentrated, leveraged bet on one narrative. The margin call exposed the lack of a safety net.

The loss is likely to reshape the conversation about AI investing. It may make investors more skeptical of funds that rely on grand narratives rather than evidence. It may also prompt regulators and allocators to ask harder questions about leverage and risk management. The fund’s collapse is a reminder that markets do not reward certainty. They reward careful, disciplined judgment.

Aschenbrenner once offered a personal hedge. A friend joked that the firm was perfectly hedged for him: either AGI happened this decade and his human capital depreciated but his financial capital grew, or no AGI happened and the firm struggled but he remained a smart person in his twenties. With the fund now reduced to a fraction of its former size, that hedge has failed in its first leg. Aschenbrenner is certainly still in his twenties.

Source: The Verge News