AI hedge funds’ assets surge; Situational Awareness tops $15B

AI-focused hedge funds are recording rapid asset growth, led by Situational Awareness. The firm, founded by former OpenAI researcher Leopold Aschenbrenner, expanded from a few hundred million dollars in seed capital to more than $15 billion in assets in less than two years.

Industry analysis shows a new generation of hedge fund managers built around artificial intelligence strategies is attracting large inflows as investors seek exposure to AI. These managers have concentrated positions in companies that provide the infrastructure for AI, including chipmakers, power suppliers, cloud providers and data-centre operators.

Situational Awareness’s concentrated holdings in AI infrastructure and semiconductors produced significant returns that accelerated asset growth and drew large institutional commitments. Other specialist managers using similar approaches have received quick inflows from pension funds, endowments and other large investors that want managers with technical expertise in AI development and deployment.

The trend is changing how new hedge funds are launched. Many recent funds are constructing portfolios focused on the hardware, software and services needed for large-scale AI models and high-volume data processing, rather than traditional sector or factor-based strategies. Investors and managers view the buildout of AI infrastructure as a multi-year investment cycle, supporting longer-term allocations to companies supplying chips, networking, cooling and cloud capacity.

Some investors warn valuations in parts of the AI ecosystem have become elevated. Despite that caution, managers with technical teams and differentiated research on AI development continue to win mandates from institutional clients.

Data analysis indicates institutional demand and early concentrated returns have helped successful AI-native managers scale quickly, attracting follow-on commitments and secondary market interest. Market activity shows AI investment extending from software into the physical and cloud systems that make large models operational, with funds prioritising semiconductors, data centres and cloud providers that meet power and compute requirements.