May 20, 2026

Real-Time Crypto Insights, News And Articles

Ex-OpenAI’s Leopold Aschenbrenner makes $13.6B AI bet tied to crypto mining firms

Leopold Aschenbrenner is doubling down on AI’s infrastructure layer, shifting capital away from semiconductor giants and into bitcoin miners and power-heavy compute providers.

The former OpenAI researcher—widely known for his warnings about geopolitical risks to advanced AI—has significantly expanded his exposure to companies that control electricity, data centers, and high-performance computing, while simultaneously building sizable bearish bets against chipmakers.

His latest 13F filing with the U.S. Securities and Exchange Commission shows disclosed equity holdings rising from $5.5 billion at the end of 2025 to $13.67 billion as of March 31.

A large portion of his long positions is concentrated in bitcoin miners and related infrastructure firms, including IREN, Core Scientific (CORZ), Riot Platforms (RIOT), CleanSpark (CLSK), Bitfarms (BITF), Bitdeer (BTDR), and Hive Digital (HIVE). These companies are increasingly repositioning themselves as providers of energy and compute capacity for AI workloads, leveraging existing access to power contracts and large-scale facilities.

As demand for AI data centers accelerates, their hybrid role as both miners and infrastructure providers has attracted growing investor attention.

Beyond mining firms, Aschenbrenner also disclosed notable stakes in Bloom Energy (BE), SanDisk (SNDK), and cloud platform CoreWeave (CRWV), reinforcing his broader thesis centered on the physical backbone of AI.

At the same time, he has taken a strong contrarian stance against the semiconductor sector. The filing reveals $7.46 billion in put options targeting chipmakers and related exchange-traded funds, including a $2.04 billion position against the VanEck Semiconductor ETF, a $1.57 billion bet against Nvidia (NVDA), and more than $1 billion in combined puts tied to Oracle (ORCL) and Broadcom (AVGO).

The positioning signals a clear strategic view: that the next phase of the AI cycle may be driven less by chip designers and more by companies that own the energy and infrastructure required to run increasingly power-intensive models.

About The Author