Greg Brockman Testifies in Musk v. OpenAI Trial
Greg Brockman, president and cofounder of OpenAI, testified for a second day in the federal trial between Elon Musk and OpenAI, according to The New York Times and Business Insider. Brockman told the court that OpenAI's compute spending has grown from $30 million in 2017 to $50 billion in 2026, Business Insider reports. He also confirmed a personal stake valued at nearly $30 billion, and Business Insider reports OpenAI is exploring an IPO. The New York Times reports Brockman described a 2017 Dota 2 victory as a triggering event after which Mr. Musk pushed to commercialize OpenAI, and Business Insider reports Musk demanded a 51% stake and engaged in heated confrontations with founders. The New York Times reports Mr. Musk is seeking $150 billion in damages in the case.
What happened
Greg Brockman, OpenAI president and cofounder, returned to the witness stand for a second day in the federal lawsuit between Elon Musk and OpenAI, according to The New York Times and Business Insider. Per Business Insider reporting of Brockman's testimony, Brockman said OpenAI's compute costs rose from $30 million in 2017 to $50 billion in 2026. Business Insider reports Brockman testified his personal stake in OpenAI is worth nearly $30 billion, and that OpenAI is exploring an initial public offering. The New York Times reports Brockman recounted a 2017 win by an OpenAI system in a Dota 2 tournament and quoted Mr. Musk as saying, "Time to make the next step for OpenAI. This is the triggering event." Business Insider reports that Brockman said Mr. Musk demanded a 51% stake in OpenAI and that negotiations led to heated confrontations. The New York Times reports Mr. Musk is seeking $150 billion in damages.
Editorial analysis - technical context
Industry observers note that large multipliers in compute spending are typical as research labs scale from research prototypes to production-grade foundation models. Rapid increases in GPU and cloud billings, dataset curation, and inference costs drive capital intensity for frontier-model work. Companies building at this scale typically face tradeoffs among engineering cost controls, model architecture choices, and accelerator procurement strategies.
Industry context
For practitioners, the trial foregrounds three cross-cutting issues: governance and legal risk around nonprofit-to-profit transitions; the visibility of personal equity stakes among technical founders; and the capital intensity of training and operating modern models. Reporting on the testimony makes those topics salient to investors, engineers, and compliance teams evaluating governance structures and budgeting for scale.
What to watch
- •Whether the jury awards damages or a remedy that affects corporate structure, as reported by The New York Times.
- •Any public disclosures or filings if OpenAI proceeds with an IPO, per Business Insider reporting.
- •Detailed breakouts of compute spending and procurement practices, which could appear in evidentiary filings or testimony and would clarify the drivers of the $50 billion figure.
Observed patterns in similar cases
Companies with fast-growing capital requirements often face intensified scrutiny over governance, valuation, and founder compensation. Legal disputes that center on mission drift and commercialization can produce lasting investor and regulatory consequences even after settlements are reached.
For practitioners
Monitor court documents and credible reporting for primary-source figures (e.g., testified compute totals and claimed valuations). Those filings, not secondary summaries alone, will matter for any operational or compliance response in adjacent organizations.
Scoring Rationale
The trial documents quantify compute and valuation figures and places governance disputes in the public record, which is notable for investors, compliance teams, and model-builders budgeting for scale. This is a major industry story with direct operational and legal relevance.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems


