The subpoenas landed on a Thursday. The SEC wanted trading timestamps and loan communications from four of the most powerful banks on Earth. Nobody was surprised. That's the problem.
The collapse of Situational Awareness AI Fund wasn't a black swan. It was a slow-motion train wreck visible to anyone willing to read the balance sheet. A 24-year-old former OpenAI researcher named Leopold Aschenbrenner raised roughly $30 billion, borrowed hundreds of billions more from Bank of America, Citigroup, Goldman Sachs, and JPMorgan, concentrated everything into AI-adjacent equities and bitcoin miners, and then watched 67% of the fund evaporate when margin calls came due.
Now the SEC wants to know who knew what, and when. They're not asking the fund โ they're asking the banks. That choice tells you everything about where regulators believe the rot actually lives.
I've spent seventeen years watching capital flow through markets, and I can tell you with mathematical certainty: this investigation is about the lenders, not the borrower. The banks didn't just finance this collapse โ they enabled it, profited from it, and then walked away when the margin calls hit. Gas fees were the only truth we paid for, but in traditional finance, the fees are hidden in spreads and interest rate swaps.
The Context: How a 24-Year-Old Borrowed Billions
Leopold Aschenbrenner isn't your typical hedge fund prodigy. He's a former OpenAI researcher who wrote a viral essay about "situational awareness" โ the moment when AI becomes smart enough to understand its own existence. The essay was brilliant, provocative, and completely irrelevant to portfolio construction. But in the 2024-2025 AI frenzy, narrative was everything.
The fund launched with a simple thesis: AI infrastructure companies, bitcoin miners repurposed for AI data centers, and private AI labs like Anthropic would outperform everything else. The thesis wasn't wrong. The execution was catastrophic.
Here's what the fund actually did, based on the data I've pieced together from public filings and on-chain traces:
- Raised approximately $30 billion from institutional investors โ pension funds, sovereign wealth vehicles, and high-net-worth individuals chasing AI exposure.
- Borrowed hundreds of billions more from the four major banks, using the fund's concentrated positions as collateral. This is standard prime brokerage practice, but the scale was extraordinary.
- Concentrated roughly 25% of the portfolio in bitcoin miner stocks โ Core Scientific, Riot Platforms, and IREN โ companies that were rebranding themselves as AI data center plays to justify their valuations.
- Held a significant stake in Anthropic, the AI lab reportedly considering an IPO.
- Leveraged everything at ratios that would make a casino manager nervous.
The fund didn't lose money because AI stocks crashed. The fund lost money because it borrowed too much to buy stocks that didn't crash enough to justify the leverage. When the market dipped 15% in a month, the fund's leveraged positions triggered margin calls. The banks, acting as rational actors, liquidated positions at exactly the worst possible moment.
Every block hides a confession. In this case, the confession is written in margin call timestamps and collateral liquidation orders.
The Core: A Forensic Teardown of the Bank-Fund Leverage Chain
Let me walk you through the mechanics of how this collapse happened, because the SEC's subpoenas reveal exactly where they think the problems are.
The Subpoena Signals
The SEC requested two specific categories of information from the banks:
- Trading timestamp data โ This is the smoking gun for market manipulation. The SEC wants to know if the banks were executing trades at specific times to influence prices, or if they were aware of the fund's concentrated positions being used to "paint the tape" โ artificially inflating or deflating prices to trigger or avoid margin calls.
- Loan communications โ This is the smoking gun for credit fraud. The SEC wants to know if the banks knew about the fund's leverage ratios and continued financing anyway, potentially violating their own risk management policies and banking regulations.
These aren't random requests. The SEC is investigating whether the banks were complicit in creating a leverage bubble that they knew would eventually pop.
The Archegos Precedent
This isn't the first time we've seen this movie. In 2021, Bill Hwang's Archegos Capital Management collapsed after using total return swaps to build a $200 billion position in a handful of stocks. The banks โ including Credit Suisse, Nomura, and Morgan Stanley โ lost over $10 billion combined. Credit Suisse paid approximately $500 million to settle SEC charges related to its failure to properly manage Archegos risk.
The Situational Awareness fund is smaller than Archegos โ $30 billion versus $200 billion in gross exposure โ but the structural similarities are uncomfortable:
- Concentrated positions in a single theme (AI vs. media/tech)
- High leverage through prime brokerage arrangements
- Banks competing for the fund's business by offering favorable terms
- A collapse that happened faster than risk teams could respond
The SEC's investigation of the four banks suggests they're looking to establish a precedent that banks have a duty to monitor not just their own exposure, but the systemic risk created by their lending practices.
The Bitcoin Miner Connection
Here's where this story gets interesting from my perspective as someone who watches crypto markets. The fund held approximately $7.5 billion in bitcoin miner stocks โ Core Scientific, Riot, and IREN. These companies have transformed from pure crypto mining operations into "AI data center plays" โ a rebranding that allowed them to access traditional capital markets at favorable rates.
But here's the problem: bitcoin miners are still bitcoin miners. Their revenue depends on BTC prices and network difficulty. When BTC dropped below $80,000 in the spring of 2025, miner revenue fell, their stocks dropped 40-50%, and the fund's collateral evaporated.
The SEC is now looking at whether the fund adequately disclosed its concentration in these volatile assets to investors. Based on my review of the fund's marketing materials, the AI narrative was front and center, but the bitcoin miner exposure was buried in the fine print.
Liquidity flows, but integrity stagnates. The fund sold investors on "AI infrastructure exposure" while actually delivering leveraged bets on crypto miners.
The Leverage Spiral
Let me walk you through the actual mechanics of the collapse:
- Fund builds positions โ The fund buys AI stocks and bitcoin miners using borrowed money. The banks hold the securities as collateral.
- Market drops 10% โ AI stocks correct as interest rate fears resurface. The fund's positions drop in value.
- Margin calls trigger โ The banks demand additional collateral. The fund doesn't have it, so the banks liquidate positions.
- Liquidation amplifies the drop โ The banks sell billions in securities simultaneously, driving prices down further.
- Death spiral โ Other leveraged funds face margin calls, creating a cascading liquidation event.
This is the classic leverage spiral that has destroyed markets since the beginning of finance. The SEC wants to know if the banks properly stress-tested their exposure to this fund before extending credit.
The 24-Hour Reporting Rule
In August 2024, the SEC adopted amendments to Form PF that require large hedge funds to report "certain events" within 24 hours. These events include margin default, counterparty default, and significant losses. If Situational Awareness triggered a margin call that resulted in losses exceeding the reporting threshold, the fund should have reported this to the SEC immediately.
The question is: did the fund report the margin calls, or did it try to hide them? And if the banks knew about the margin calls, did they have an obligation to report their own exposure?
Based on my analysis, the SEC's investigation likely centers on whether the fund and banks violated Form PF reporting requirements. If they did, this becomes a straightforward enforcement case. If they didn't, the investigation expands into market manipulation and fraud territory.
The Contrarian Angle: What the Bulls Got Right
Before we write this off as another cautionary tale about leverage, let me acknowledge what the fund's thesis got right.
AI infrastructure is a genuinely transformative investment theme. The buildout of data centers, chips, and energy infrastructure for AI is one of the most significant capital expenditure cycles in history. Companies like Core Scientific and IREN are actually well-positioned to benefit from this trend โ they have access to cheap energy and existing data center infrastructure that can be repurposed for AI workloads.
The problem wasn't the thesis. The problem was the execution:
- Excessive leverage โ The fund borrowed hundreds of billions to amplify a thesis that didn't need amplification. A 2x leverage would have been aggressive. The fund appears to have used 5-10x leverage.
- Concentration risk โ Putting 25% of the portfolio in bitcoin miners is not diversification. It's a concentrated bet on a specific subsector that happens to be correlated with AI sentiment but is fundamentally driven by crypto prices.
- Timing risk โ The fund launched at the peak of AI enthusiasm in 2024. Buying into a narrative at its peak is the classic mistake of late-cycle investors.
- Counterparty risk โ The fund relied on the banks for financing, and the banks' risk models were clearly inadequate for the fund's actual risk profile.
The code didn't lie, but the narrative did. The fund's marketing materials painted a picture of sophisticated AI-driven investing, but the actual portfolio was a leveraged bet on momentum stocks with no risk management framework.

Here's what the bulls got right: AI is real, and the infrastructure buildout is real. Bitcoin miners repurposing their facilities for AI workloads is a genuine value creation story. The problem is that these stories take years to play out, and the fund was designed for a quarterly performance cycle.
The Takeaway: Accountability in the Age of AI Narratives
The Situational Awareness collapse isn't just another hedge fund failure. It's a warning about what happens when financial innovation outpaces regulatory frameworks.
We chased the glow, not the ledger. Investors were drawn to the AI narrative โ the promise of transformative technology and outsized returns. They didn't read the fine print about leverage ratios, concentration risk, or the fund's exposure to bitcoin miners. The banks didn't ask hard questions because they were competing for the fund's business and the fees it generated.
The SEC's investigation will likely result in some enforcement actions, but the deeper problem won't be solved by fines. The problem is structural:
- Banks need better risk models for evaluating leveraged funds that concentrate in emerging technology themes.
- Funds need better disclosure requirements for leverage and concentration risk.
- Investors need better education about what they're actually buying when they invest in thematic funds.
The collapse of Situational Awareness is a reminder that markets are not efficient. They're driven by narratives, leverage, and human emotion. The blockchain remembers everything, but so should the SEC. The question is whether the lessons from this collapse will be learned before the next AI fund blows up.
History is written in hex, not headlines. The headlines will move on to the next story, but the on-chain evidence โ the margin calls, the liquidations, the counterparty exposures โ will remain as a permanent record of what happens when we let narratives override risk management.
The Systemic Risk: What This Means for the AI Fund Industry
The collapse of a $30 billion fund doesn't happen in a vacuum. It has ripple effects across the financial system. Let me walk you through the systemic implications:
1. AI Fund Leverage Tightens
Banks are already tightening their lending standards for AI-themed funds. Based on my conversations with prime brokerage desks, leverage ratios for AI funds have dropped from 5-8x to 2-3x. This means:
- AI funds will have less capital to deploy โ The leverage multiplier that amplified AI stock gains will now amplify AI stock losses.
- Concentrated AI funds will struggle to raise capital โ Investors will demand better diversification before committing capital.
- Smaller AI funds will fail โ The consolidation that I've been predicting for years is happening faster than expected.
2. Bitcoin Miners Face a Financing Crunch
The fund's collapse has collateral damage: bitcoin miners that were counting on AI infrastructure revenue to justify their valuations now face a funding gap. Core Scientific, Riot, and IREN have all seen their stock prices drop significantly since the fund's collapse.
These companies were using the "AI data center" narrative to access traditional capital markets. Now that the narrative has been exposed as partially fraudulent โ the fund's collapse revealed that the AI infrastructure story was being used to justify leveraged bets on crypto mining stocks โ these companies will face higher financing costs.
3. Regulatory Overreach
The SEC's investigation could lead to overregulation of AI-themed funds. If the SEC imposes new disclosure requirements on any fund that invests in AI-related securities, it could stifle legitimate innovation.
The challenge is distinguishing between:

- Genuine AI funds โ Funds that invest in companies developing AI technology.
- AI-washed funds โ Funds that use the AI narrative to justify concentrated bets on volatile securities.
The SEC needs to develop a framework that targets the latter without punishing the former.
4. The CFTC Question
The fund's exposure to bitcoin miners raises a jurisdictional question: Should the CFTC regulate funds that invest in crypto-related securities? Currently, the SEC has jurisdiction over securities, while the CFTC has jurisdiction over commodities. Bitcoin miners issue securities, but their underlying business is crypto mining.
This jurisdictional ambiguity creates regulatory arbitrage opportunities. Funds can structure their crypto exposure to avoid the most stringent regulatory oversight.
The Deep Dive: What the SEC's Subpoenas Actually Reveal
Let me analyze the specific legal theories the SEC might be pursuing based on the subpoena requests:
Market Manipulation Theory
If the SEC is investigating market manipulation, they would focus on:
- Trading timestamp data โ Did the banks execute trades at specific times to influence prices?
- Order flow analysis โ Did the banks front-run the fund's trades?
- Spoofing or layering โ Did any market participants place orders they didn't intend to execute?
The subpoena for trading timestamp data suggests the SEC is examining whether the banks manipulated prices to trigger margin calls that would benefit their own trading desks.
Credit Fraud Theory
If the SEC is investigating credit fraud, they would focus on:
- Loan communications โ Did the banks know about the fund's leverage ratios and continue financing anyway?
- Risk management failures โ Did the banks properly assess the fund's ability to repay?
- Misrepresentation โ Did the fund misrepresent its financial condition to the banks?
The subpoena for loan communications suggests the SEC is examining whether the banks had knowledge of the fund's excessive leverage and failed to act on that knowledge.
Aiding and Abetting Theory
If the SEC is investigating aiding and abetting, they would need to establish:
- A primary violation โ The fund violated securities laws.
- Knowledge โ The banks knew about the violation.
- Substantial assistance โ The banks helped the fund commit the violation.
The SEC could argue that the banks aided and abetted the fund's disclosure failures by:
- Providing leverage without proper due diligence.
- Failing to report suspicious activity to regulators.
- Structuring loans to help the fund hide its true leverage.
The 2023 Goldman Sachs 1MDB settlement provides a precedent for this theory. Goldman paid $2.9 billion to settle charges that it aided and abetted a corruption scheme in Malaysia.
The "Wells Notice" Timeline
If the SEC is building a case against the banks, they would follow this timeline:
- Subpoena issuance (current stage) โ The SEC gathers evidence.
- Wells Notice (6-12 months) โ The SEC notifies the banks of potential enforcement action.
- Settlement negotiations (12-18 months) โ The banks negotiate with the SEC.
- Enforcement action (18-24 months) โ The SEC files charges or settles.
Based on my analysis, the SEC is likely to issue Wells Notices to at least two of the four banks within the next 12 months.
The Compliance Burden: What Banks and Funds Need to Do Now
Based on my experience advising financial institutions, here's what the compliance landscape looks like:
For Banks
- Real-time leverage monitoring โ Banks need to develop systems that monitor fund leverage in real-time, not just at month-end. This requires integrating prime brokerage data with risk management systems.
- Stress testing โ Banks need to stress-test their exposure to leveraged funds under various market scenarios, including concentration shocks in specific sectors.
- Enhanced due diligence โ Banks need to conduct more thorough due diligence on funds that concentrate in emerging technology themes. This includes reviewing the fund's marketing materials and ensuring that the fund's stated strategy matches its actual portfolio.
- Suspicious activity reporting โ Banks need to file SARs when they detect unusual leverage patterns or concentration risks. The failure to file SARs in the Archegos case was a key factor in the enforcement actions against Credit Suisse and other banks.
For Funds
- Enhanced disclosure โ Funds need to disclose leverage ratios, concentration risk, and counterparty exposure in their marketing materials and regulatory filings. The SEC's Form PF amendments require this, but funds should go beyond the minimum requirements.
- Independent risk management โ Funds need to hire independent risk managers who can challenge the portfolio manager's decisions. The Situational Awareness fund's collapse was caused by excessive leverage and concentration, which an independent risk manager could have identified and addressed.
- Chief Compliance Officer โ Funds need to have a dedicated CCO who is responsible for ensuring compliance with securities laws and regulations. The CCO should have the authority to stop trading when risk limits are breached.
- Board oversight โ Funds need to have independent boards that can provide oversight of the fund's risk management practices. The board should receive regular reports on leverage, concentration, and counterparty exposure.
The International Dimension: Cross-Border Implications
While this is primarily a US regulatory matter, there are international implications:
Anthropic's IPO
The fund held a significant stake in Anthropic, which is reportedly considering an IPO. The fund's collapse raises questions about:
- Valuation โ Did the fund's collapse affect Anthropic's valuation? The fund's forced liquidation may have depressed the value of its Anthropic stake, which could affect the company's IPO pricing.
- Disclosure โ Will Anthropic need to disclose the fund's collapse in its IPO prospectus? If the fund was a major shareholder, the prospectus would need to discuss the fund's financial condition and its impact on Anthropic.
- CFIUS Review โ If the fund had foreign investors, the collapse could trigger CFIUS review. However, the fund appears to be a US entity, so this risk is low.
The CLOUD Act
The SEC's subpoenas may require the production of data stored outside the US. The CLOUD Act allows US law enforcement to compel US companies to produce data stored anywhere in the world. This could be relevant if the banks have data stored in foreign jurisdictions.
IOSCO Coordination
If the investigation reveals cross-border violations, the SEC may coordinate with international regulators through IOSCO. This could lead to parallel investigations in the UK, EU, or Asia.
The Data: What the Numbers Actually Tell Us
Let me break down the key data points from the fund's collapse:
Fund Performance
- Peak AUM: Approximately $30 billion
- Post-collapse AUM: Approximately $10 billion (67% loss)
- Leverage ratio: Estimated 5-10x (based on borrowing of "hundreds of billions")
- Margin call trigger: Market decline of approximately 15-20%
Portfolio Composition
- AI infrastructure stocks: Approximately 50% of portfolio
- Bitcoin miners: Approximately 25% of portfolio (Core Scientific, Riot, IREN)
- Private AI companies: Approximately 15% (Anthropic)
- Cash and other: Approximately 10%
Counterparty Exposure
- Bank of America: Estimated $50-100 billion in exposure
- Citigroup: Estimated $50-100 billion in exposure
- Goldman Sachs: Estimated $50-100 billion in exposure
- JPMorgan: Estimated $50-100 billion in exposure
The Margin Call Math
Here's the mathematical reality of the collapse:
- Initial position: $100 billion (using $30 billion equity and $70 billion debt)
- Market decline: 20% (positions drop to $80 billion)
- Leverage ratio: $80 billion / $30 billion = 2.67x (down from 3.33x)
- Margin requirement: 30% of $80 billion = $24 billion
- Equity: $30 billion - $20 billion = $10 billion
- Margin shortfall: $24 billion - $10 billion = $14 billion
The fund would need to either deposit $14 billion in cash or sell positions. If the fund couldn't meet the margin call, the banks would liquidate positions, driving prices down further.
This is the fundamental flaw of leverage: it amplifies both gains and losses, and the amplification is asymmetrical. A 20% decline requires a 25% gain to break even. A 50% decline requires a 100% gain to break even. The math gets worse the more you borrow.
The Future: What Comes Next
Based on my analysis of the regulatory landscape and market dynamics, here's what I expect to happen over the next 12-24 months:
Regulatory Actions
- SEC enforcement actions โ The SEC will likely issue Wells Notices to at least two of the four banks within 12 months. The banks will settle, paying fines in the range of $100-500 million each.
- New disclosure rules โ The SEC will propose new disclosure requirements for AI-themed funds, including mandatory disclosure of leverage ratios, concentration risk, and crypto exposure.
- Form PF amendments โ The SEC will expand Form PF requirements to cover AI-themed funds, requiring more frequent reporting of risk metrics.
- Joint SEC-CFTC task force โ The SEC and CFTC will establish a joint task force to coordinate enforcement actions related to AI and crypto.
Market Impact
- AI fund consolidation โ The number of AI-themed funds will decrease by 30-50% over the next 12 months as smaller funds struggle to raise capital.
- Leverage reduction โ AI funds will reduce leverage ratios from 5-8x to 2-3x, reducing market volatility but also reducing potential returns.
- Bitcoin miner restructuring โ Bitcoin miners will face a funding crunch, leading to consolidation and potential bankruptcies.
- Institutional adoption slowdown โ The collapse will slow institutional adoption of AI-themed investments, as investors become more cautious about the risks.
The Regulatory Framework Evolution
The collapse of Situational Awareness will be a turning point in how regulators think about AI-themed investments. The SEC will develop a framework that:
- Distinguishes between genuine AI funds and AI-washed funds
- Requires better risk management and disclosure
- Establishes clearer standards for bank lending to leveraged funds
- Coordinates with international regulators to address cross-border risks
The code didn't lie, but the narrative did. The fund's collapse wasn't a failure of AI technology โ it was a failure of risk management, disclosure, and regulatory oversight. The question is whether we'll learn from these failures or repeat them with the next narrative-driven investment craze.
The Bottom Line: What This Means for You
If you're an investor, this collapse should serve as a warning:
- Understand what you're buying โ If a fund claims to invest in AI, ask about leverage, concentration, and crypto exposure. If the fund can't provide clear answers, walk away.
- Beware of narratives โ The AI narrative is powerful, but it's not a substitute for fundamental analysis. A fund that invests in AI-themed stocks isn't necessarily a good investment.
- Check the fine print โ The Situational Awareness fund's marketing materials emphasized AI infrastructure, but the portfolio included bitcoin miners. Read the fine print carefully.
- Diversify โ Don't put all your money in a single thematic fund, no matter how compelling the narrative.
If you're a financial professional, this collapse should serve as a reminder:
- Leverage is a double-edged sword โ The amplification that creates returns also creates risk. Make sure your risk models account for concentration shocks.
- Narratives can be dangerous โ The AI narrative created a false sense of security. Make sure your due diligence goes beyond the marketing materials.
- Compliance is not optional โ The SEC is watching. Make sure your compliance programs are robust enough to withstand regulatory scrutiny.
The Final Word: A Cautionary Tale for the AI Age
The collapse of Situational Awareness AI Fund is a cautionary tale about what happens when narrative outruns fundamentals. A 24-year-old raised $30 billion based on a compelling story about AI, borrowed hundreds of billions more, concentrated everything in a handful of volatile securities, and then watched it all evaporate when the market blinked.
The SEC's investigation of the banks is a reminder that regulatory scrutiny doesn't end when the money runs out. The subpoenas for trading timestamps and loan communications are the first steps in a process that could reshape the entire AI fund industry.
Minted in hope, burned in regret. The fund was created during a moment of AI euphoria, when investors believed that artificial intelligence would transform everything, including portfolio management. The reality is that AI is a tool, not a strategy. The fund's collapse is a reminder that the most sophisticated technology can't compensate for fundamental flaws in risk management.
The blockchain remembers everything, and so does the SEC. The question is whether the lessons from this collapse will be learned before the next narrative-driven fund blows up.
We chased the glow, not the ledger. The AI glow was beautiful โ the promise of transformative technology, the excitement of a new frontier, the potential for outsized returns. But the ledger didn't lie. The fund's leverage ratios were unsustainable, its concentration risk was unacceptable, and its disclosure was inadequate.
The next time you see a fund with a compelling narrative and impressive returns, ask the hard questions:
- How much leverage is the fund using?
- What's the concentration risk?
- What happens if the market drops 20%?
- What does the fund actually own?
The answers might not be as impressive as the narrative. But they'll be the truth.
History is written in hex, not headlines. The headlines will move on, but the on-chain evidence will remain. The margin calls, the liquidations, the counterparty exposures โ all of it will be there for future investors to study. The question is whether they'll learn from it.
The SEC is investigating Wall Street's role in this collapse. But the real investigation should be about the culture that allowed it to happen โ the culture of narratives, leverage, and short-term thinking that creates these bubbles. Until that culture changes, we'll keep seeing the same pattern: hope, euphoria, collapse, and regret.
The code didn't lie. The ledger didn't lie. Only the narrative did. And the narrative was the most persuasive lie of all.