Sam Altman says OpenAI has $20B ARR and about $1.4 trillion in data center commitments (techcrunch.com)

OpenAI CEO Sam Altman publicly clarified the company’s scale and plans, saying OpenAI expects to finish the year above $20 billion in annualized revenue run rate and is “looking at commitments of about $1.4 trillion over the next 8 years.” The comment, framed as a response to controversy over off‑hand CFO remarks about government‑backed loans, also previewed new revenue channels: an enterprise offering (OpenAI already claims ~1 million business customers), a nascent “OpenAI for Science” push, and the possibility of directly selling compute capacity as an “AI cloud.” Technically and strategically, these figures signal enormous projected demand for training and inference infrastructure and a potential move by OpenAI from a pure software/model provider toward vertical integration in data‑center operations and cloud compute. Committing up to $1.4T implies major capital spending, supply‑chain, energy and regional deployment considerations, and increases competitive pressure on hyperscalers (AWS/GCP/Azure) if OpenAI sells compute. Financing (equity or loans) and the fact OpenAI doesn’t yet own a broad data‑center network add uncertainty, but the announcement underscores how rapidly AI economics are scaling—pushing model size, throughput, and infrastructure investment to new heights with material implications for pricing, access, and industry consolidation.