On August 15, 2026, the SEC received a 13F filing from Situational Awareness LP. The numbers were not subtle. Micron and SanDisk—two storage chip stocks—accounted for 55% of the fund's disclosed equity portfolio. By the end of Q2, that was roughly $11.2 billion in two names. The rest of the portfolio was a chain of AI infrastructure bets: Bloom Energy, TSMC, Nebius, CoreWeave, Core Scientific, and a handful of mining and data-center operators. Put options that had hedged against semiconductor exposure in Q1 were slashed. The fund was all-in on the AI hardware narrative.
This is not a DAO treasury. It is a hedge fund run by Leopold Aschenbrenner, a former OpenAI researcher who built his reputation on predicting the trajectory of artificial intelligence. But the structural risk pattern is identical to what I have audited in dozens of decentralized autonomous organizations during the 2022 crash. The same absence of diversification. The same disregard for correlation. The same assumption that the story will continue.
Context: The Fund and the Filing Situational Awareness LP first appeared in the 13F system in Q1 2026 with a mixed long-short strategy. The fund held significant put options on SMH, NVIDIA, Broadcom, AMD, Oracle, Micron, and TSMC. That was a prudent hedge against a crowded trade. By Q2, those puts were largely gone. Instead, the fund doubled down on the long side, concentrating capital into the very names it had been hedging against. The filing, which reports holdings as of June 30, 2026, shows Micron surging from $5.86 million to $5.574 billion. SanDisk went from $724 million to $5.674 billion. The two stocks alone represented more than half the portfolio. Bloom Energy jumped to $1.899 billion. TSMC to $1.265 billion. A new position in Nebius, a European AI cloud provider, reached $1.233 billion. CoreWeave, Core Scientific, Applied Digital, IREN, Riot—all added or increased.
The shift from hedging to concentrated long is a governance decision. It reflects a conviction that the AI hardware cycle is not just real but accelerating. But conviction is not a risk management framework. In my work with DAO treasuries, I have seen the same logic: a governance committee becomes convinced of a thesis, votes to allocate 80% of assets to a single protocol, and then watches the portfolio collapse when the thesis fails. The difference is that DAOs have built-in transparency and often require multi-sig approvals. Situational Awareness LP is a traditional fund, so its investors must rely on the GP's judgment. The 13F is the only public signal, and it is quarterly and lagged.
Core: The Structural Risk of 55% Let me be precise. A portfolio with 55% in two stocks that are highly correlated—both semiconductor memory manufacturers, both exposed to the same AI demand cycle, both subject to the same geopolitical risks (Taiwan, export controls, commodity pricing)—is not an investment. It is a leveraged bet on a single narrative. The fund's other positions only amplify the correlation. Bloom Energy provides fuel cells for data centers. TSMC is the foundry for most AI chips. Nebius, CoreWeave, Core Scientific, Applied Digital, IREN, Riot are all infrastructure plays: compute, power, mining. The entire portfolio is a single vector: AI hardware demand. If the demand narrative falters, every position moves in the same direction.
From my audit experience during the 2022 crash, I can tell you exactly what happens next. In a bull market, correlation is low because everything rises. In a correction, correlation goes to 1.0. All these stocks will fall together. The fund's total equity exposure, as disclosed, is roughly $20 billion in public equities. If the AI hardware sector drops 30%, the portfolio loses $6 billion. But the real risk is leverage. The 13F does not disclose borrowing, margin accounts, or derivatives beyond the puts that were already removed. If the fund used margin to amplify these positions—and the size of the quarter-over-quarter increase suggests leverage was involved—then a 30% decline could trigger margin calls, forced liquidations, and a cascade that takes down not just the fund but the stocks themselves.
Trust the code, but verify the architecture. The architecture of this portfolio is fragile. The fund's governance structure allowed a single individual to concentrate risk to a degree that would be unacceptable in any DAO I have audited. Most DAOs I work with have treasury diversification policies that limit any single asset to 10-15% of the portfolio. Even the most aggressive ones cap at 20%. Situational Awareness LP exceeded that by a factor of nearly three. This is not a criticism of the investment thesis. The thesis might be correct. But the structure is indefensible.
Contrarian: The Rebound Does Not Validate the Strategy A skeptic might point to the recent market action. After a July sell-off that hit Micron, SanDisk, and the entire semiconductor chain, the sector rebounded in early August. Inflation data cooled, AI earnings sentiment recovered, and stocks like SanDisk, Micron, and CoreWeave surged. The Philadelphia Semiconductor Index erased its monthly decline. The fund's portfolio, which was underwater in July, likely recovered much of the loss. A casual observer might say the strategy worked.
That is a dangerous fallacy. The rebound does not validate the concentration. It only masks the risk. The portfolio was one bad CPI print away from a liquidity crisis. If July had been a 40% drawdown instead of a 20% dip, the margin calls would have forced liquidations into a falling market, amplifying the losses and potentially triggering a systemic event. The fact that it did not happen this time does not mean the structure is sound. In my experience, a governance failure is not defined by the outcome. It is defined by the decision process. The decision to concentrate 55% of a portfolio into two stocks without a corresponding hedge is a governance failure regardless of what happens next.
Governance is not a feature; it is the foundation. The foundation of this fund is sand. The 13F filing is a quarterly snapshot, but it reveals a pattern of escalating risk. In Q1, the fund had hedges. By Q2, the hedges were gone. The GP made a deliberate choice to remove the safety net. That is not conviction. That is recklessness. In the DAO world, we would call this a governance attack from within—a single actor with too much power over the treasury. The only difference is that here, the actor is the GP, and the investors signed up for it.
Takeaway: The Ledger Remembers What the Community Forgets The ledger of this fund's 13F filings will be studied by future analysts as a case study in risk concentration. The question is whether investors will demand better governance. In the crypto space, we have learned that transparency is not enough. You need enforceable rules, diversification mandates, and emergency protocols. Situational Awareness LP has none of these. The fund's performance will depend on the AI narrative continuing uninterrupted. But narratives do not respect portfolios. The market will remind us that efficiency without oversight is just faster risk.
I will be watching the next 13F filing. If the fund is still alive, I will check whether the portfolio has been restructured. If it has not, then the lesson is clear: traditional finance still has not learned what decentralized governance has been screaming for years. Trust the code, but verify the architecture. And if the architecture is a single point of failure, do not be surprised when it fails.
The ledger remembers what the community forgets. In this case, the community is the fund's investors, and the memory will be written in red.