Musk v. Altman: Greg Brockman said OpenAI expects to spend $50B on computing in 2026, up from ~$30M in 2017; OpenAI told investors it would spend ~$600B by 2030
Rachel Metz /Bloomberg:
Context & Ripple Effects
Brockman’s testimony puts OpenAI’s computing plans alongside evidence of the company’s unusually large capital commitments: a Microsoft executive later testified that Microsoft had spent more than $100B on OpenAI through investments and infrastructure.
The same court coverage has focused on the value of OpenAI insiders’ stakes and the organization’s governance disputes, making the compute-spending disclosures relevant to both OpenAI’s financing needs and the economic stakes behind its structure.
First-order effects
- OpenAI is planning a far larger near-term compute budget, making access to capital and infrastructure central to its product organization under Brockman.
- The reported spending trajectory gives investors a much bigger funding requirement to assess through 2030, while increasing the importance of partners that can finance or build infrastructure.
Second-order effects
- Microsoft’s already substantial OpenAI investment and infrastructure spending becomes more strategically consequential, because OpenAI’s expansion depends on continued capacity at a scale far beyond its earlier operations.
- Other AI developers and infrastructure providers face a higher bar for competing for compute capacity and financing as OpenAI commits to a larger share of the available capital pool.
Third-order effects
- If this spending pattern persists, leading AI development may become increasingly concentrated among organizations able to combine frontier-model talent with exceptionally large, long-duration infrastructure financing.
- The gap between the commercial value of insider stakes and the funding required to operate at scale is likely to keep governance and control arrangements under pressure, particularly for organizations with nonprofit-linked structures.
The trend: Frontier AI is shifting from a software-led race into a capital-intensive infrastructure contest in which compute access and financing shape competitive power.