Situational Awareness LP, the hedge fund built around Leopold Aschenbrenner's 2024 Situational Awareness essay, reportedly grew to roughly $20 billion in assets before being forced to unwind after steep losses, according to reporting from Disruption Banking and TheSiliconReview. According to Disruption Banking, which cites Financial Times reporting based on an investor letter dated July 24, Aschenbrenner reported a 439% net return for the fund in the first half of 2026 and acknowledged the fund had "not been immune" to recent market swings, "particularly in Asia." Disruption Banking's account of that FT reporting also said the fund had used borrowing to magnify its returns, a strategy that can amplify losses when markets reverse.
TheSiliconReview, citing "multiple sources with direct knowledge of the situation," reported the unwind in more detail. One source familiar with the situation told the outlet:
"The losses were significant enough that his brokers required him to unwind everything."
The same report said Aschenbrenner's losses were "in the hundreds of millions of dollars, though the exact figure could not be confirmed" — TheSiliconReview's own hedge, which this article preserves. A separately reported $45 billion figure could not be verified by PutCall and does not appear here. Aschenbrenner has not publicly commented, per TheSiliconReview.
Who Is Leopold Aschenbrenner?
According to Disruption Banking, Aschenbrenner joined OpenAI's Superalignment team in 2023 and was fired in April 2024; by late that year he had launched Situational Awareness LP as CIO and managing partner. Early investors reportedly included Stripe co-founders Patrick and John Collison, former GitHub CEO Nat Friedman, and investor Daniel Gross, with Jane Street later joining — notable, Disruption Banking noted, "given how rarely the firm backs external managers." He published the essay Situational Awareness in June 2024, arguing that continued scaling of AI compute made artificial general intelligence plausible by around 2027.
What the Fund Actually Held
Disruption Banking reported the fund's public holdings "have included names such as Riot Platforms, Core Scientific, CleanSpark, Bloom Energy, AMD and Oracle" — several of which "contributed to the strong first-half performance," with the outlet adding that "some of them are the reason why Situational Awareness is seeking to raise new capital" now. The portfolio mixed AI infrastructure, cloud, energy and crypto-mining companies rather than representing a pure semiconductor portfolio — of the six, only AMD is a chipmaker in the conventional sense. None of these six tickers currently rank among PutCall's top WallStreetBets names.
WSB's Version of the Same Trade
What connects this to PutCall isn't a shared ticker list — it's a shared theme. The fund was a leveraged, concentrated bet on the AI buildout; WallStreetBets is currently concentrated on a different AI-adjacent name in large volume. Micron (MU) led all WallStreetBets discussions tracked by PutCall as of Aug. 1, with 107 mentions and a Neutral sentiment score of 60/100. SanDisk (SNDK), another memory and semiconductor name, also ranked in PutCall's top 10 with 42 mentions, Neutral at 56. Mention counts reflect discussion volume, not investment recommendations or expected stock performance. Retail traders concentrating on one or two names also face concentration risk, even if the circumstances differ substantially from those of a leveraged institutional fund.
PutCall continuously tracks WallStreetBets mentions, sentiment, and SEC Form 4 insider filings for these tickers, free, at putcall.nl.
What the Insider Trading Filings Show
One of PutCall's standard cross-checks is whether corporate insiders are trading alongside — or against — retail sentiment. As of this week, there are no fresh SEC Form 4 insider trading filings logged for either MU or SNDK on PutCall's insider tracker. The absence of insider buying or selling isn't inherently bullish or bearish — it simply means there isn't additional insider activity reinforcing the current retail narrative. If insiders begin buying during a semiconductor pullback, that would provide an additional data point beyond retail sentiment alone.
The Retail Lesson, Without the Leverage
The reporting from Disruption Banking and TheSiliconReview attributes the fund's reported unwind primarily to leverage, not to Aschenbrenner's AI thesis being wrong — the 439% first-half return, if accurate, argues the opposite. Most retail traders following WallStreetBets discussions into MU or SNDK this week aren't using that kind of leverage or facing broker-mandated liquidation, which removes the specific mechanism reported in Situational Awareness's unwind. Leverage can accelerate losses, but concentration alone is enough to make any single-stock trade far riskier than a diversified portfolio. That's the broader lesson retail investors can take from this week's AI momentum trade.
This article relies on reporting from Disruption Banking and TheSiliconReview, together with PutCall's own market data. References to Financial Times reporting are attributed through Disruption Banking.
Sources
- Disruption Banking — "Can the Situational Awareness Hedge Fund Raise Capital After its 439% H1 Gain?" (Andrew Samu, July 30, 2026)
- TheSiliconReview — "AI Investor Leopold Aschenbrenner Forced to Unwind All Public Stock Positions after Steep Losses, Sources Say" (Sashindra Suresh, July 31, 2026)
- putcall.nl/api/dashboard — WallStreetBets mention counts and sentiment scores, PutCall first-party data
- putcall.nl/api/insider-alerts — SEC Form 4 insider trading filings, PutCall first-party data
References to Financial Times reporting are relayed through Disruption Banking's coverage, not read directly by PutCall.