Aschenbrenner 13F Shows AI and Industrials Over Crypto
Leopold Aschenbrenner’s latest quarterly disclosure has drawn attention for what it does not show as much as what it does. The hedge fund manager’s reported U.S. equity book was concentrated in AI infrastructure, power, data centers and industrial names, while crypto-related exposure was limited to Bitcoin mining and adjacent compute infrastructure. That matters now because the filing reinforces a broader market split: capital is still flowing toward the physical buildout behind artificial intelligence, even as digital-asset equities remain a smaller part of the trade.
Overview
- Large disclosed equity book: Situational Awareness LP reported roughly $5.52 billion in U.S. equity holdings across 29 positions in its Q4 2025 13F, indicating a highly concentrated portfolio [3][4].
- AI infrastructure focus: Top disclosed holdings included CoreWeave, Bloom Energy, Intel, Lumentum and Core Scientific, pointing to a preference for compute, power and data-center capacity [2][3].
- Bitcoin miner exposure: The fund added or expanded positions in IREN, Cipher Mining, Riot Platforms, Bitdeer and Applied Digital, linking crypto exposure to infrastructure rather than spot digital assets [2][4].
- Core Scientific stake disclosed: An amended Schedule 13D showed a 9.4% stake in Core Scientific, or 28,756,478 shares, giving the fund a larger voice in a company straddling Bitcoin mining and AI hosting [2][4].
- Crypto was not the main theme: The public filing narrative centered on AI and industrial capacity, with crypto appearing mainly through mining firms that are increasingly repositioning for high-performance computing [2][3].
- Important limitation: A 13F captures only certain U.S. equity and option positions, so it does not provide a full picture of total exposure across derivatives, cash, private assets or non-U.S. holdings [1].
Subscribe to our Social Media for Exclusive Crypto News and Insights 24/7!
Aschenbrenner 13F and the AI trade
The filing has become a reference point for investors tracking where large, early-stage capital is being deployed in the AI cycle. Aschenbrenner, a former OpenAI researcher, built his public book around the infrastructure that supports artificial intelligence rather than the software layer that consumes it [2][3]. That included power, data-center operators and chip-adjacent names, a mix that market participants view as a direct bet on the bottlenecks of AI expansion [2][5].
The scale of the disclosed positions was notable. One report cited about $5.52 billion in U.S. equity exposures across 29 holdings, up sharply from much smaller reported assets earlier in the year [2][4]. Another account described the public holdings as a barbell around compute, energy and industrial capacity [1][5]. Interpretation based on available data: the portfolio suggests conviction that the market is still underpricing the companies that supply electricity, cooling and physical infrastructure for AI workloads.
Disclosed holdings versus theme
| Disclosed theme | Representative holdings | Market implication |
|---|---|---|
| AI compute and infrastructure | CoreWeave, Intel, Lumentum | Exposure to the buildout behind training and inference demand [2][3] |
| Power and industrial capacity | Bloom Energy, Applied Digital | A bet on electricity supply and data-center support assets [2][3] |
| Bitcoin mining / HPC transition | Core Scientific, IREN, Cipher Mining, Riot, Bitdeer | Crypto-linked exposure framed through infrastructure and hosting capacity [2][4] |
Crypto appears as a secondary sleeve
The key point for digital-asset investors is that crypto did not emerge as the center of gravity in the filing. Instead, the public positions tied to crypto were mostly in miners and infrastructure operators that are trying to diversify beyond Bitcoin production. Core Scientific was singled out because of both its share count and the reported 9.4% stake, while other miner names reflected a broader view that mining companies can monetize excess power and server capacity in AI hosting [2][4].
That distinction matters for market structure. Analysts note that the public market’s preferred crypto equity trade is no longer just hashrate exposure. It is increasingly a hybrid play on power access, rack space and data-center utilization [2][3]. For investors, that broadens the opportunity set, but it also dilutes the clean Bitcoin-beta narrative that once defined listed miners.
Crypto-related positions in the filing
| Company | Reported role in portfolio | Why it matters |
|---|---|---|
| Core Scientific | Largest disclosed crypto-linked thesis through a 9.4% stake | Signals conviction in a miner’s pivot toward AI hosting [2][4] |
| IREN | Bitcoin miner with growing compute ambitions | Reflects the market’s re-rating of energy-rich infrastructure [2][4] |
| Cipher Mining | Miner with data-center relevance | Highlights investor interest in power and facility scale [2][4] |
| Riot Platforms | Established miner | Shows continued exposure to public mining equities despite volatility [2][4] |
| Bitdeer | Miner and infrastructure name | Suggests preference for firms with optionality beyond pure mining [2][4] |
What the filing says about investor behavior
For crypto traders, the more important signal may be what was left out. There was no evidence in the public filing of a major direct allocation to spot Bitcoin, Ether or broad digital-asset funds. The disclosed book instead favored companies that can benefit from AI capex, industrial electrification and compute scarcity [2][3]. That is consistent with a market that still rewards cash-flow visibility, physical assets and power access over more speculative crypto adjacencies.
Market participants view that as a risk to pure-play miners and a tailwind for miners that can successfully convert excess infrastructure into AI hosting revenue. It also creates a more selective market. Investors appear willing to back the category, but only when the business model extends beyond block rewards [2][4].
Risks and limits in the read-through
The main limitation is that a 13F is incomplete. It excludes many forms of exposure, including non-U.S. holdings, OTC swaps and cash-like positions [1]. That means the filing cannot be treated as a full map of Aschenbrenner’s total risk. It is a snapshot of selected U.S. reported positions only.
There is also execution risk in the thesis itself. If AI demand slows, capital expenditure moderates or power markets tighten, the infrastructure names that have benefited most could re-rate quickly. For crypto miners, the downside is sharper: companies that fail to convert mining assets into broader compute utility may remain exposed to Bitcoin price swings and margin pressure. Interpretation based on available data: the trade works best if AI buildout stays strong and if miners can keep monetizing their power and real estate outside of Bitcoin production.
The filing leaves a clear takeaway for the market. Aschenbrenner’s public book still points investors toward the industrial and infrastructure layers of the AI economy, while crypto remains visible mainly through the subset of miners that can evolve into compute platforms. That keeps Bitcoin-related equities in play, but as an adjacent trade rather than the lead story.
Sources
- https://www.kucoin.com/blog/en-leopold-aschenbrenner-full-q1-2026-13f-breakdown
- https://www.mexc.com/news/850499
- https://www.tradingview.com/news/cointelegraph:9de303d84094b:0-ex-openai-researcher’s-hedge-fund-reveals-big-bitcoin-miner-bets-in-new-sec-filing/
- https://www.linkedin.com/posts/stephaniesoquet_situational-awareness-13f-shows-55b-bet-activity-7434955933585313792-kkco
- https://linas.substack.com/p/leopold-aschenbrenner-situational-awareness-fund-portfolio-playbook
