According to reporting by CNBC, OpenAI's annualized revenue is running about $20 billion below figures the company had previously signaled to the market. The downward revision comes as OpenAI has announced a string of massive infrastructure deals in recent months with compute providers including Nvidia, Oracle and CoreWeave, collectively worth hundreds of billions of dollars over multiple years.
The gap between actual revenue and earlier projections raises questions about OpenAI's ability to fund these long-term commitments through operating income alone. The company, which remains unprofitable despite rapid growth in ChatGPT subscriptions and API usage, continues to rely heavily on successive funding rounds and support from strategic partners such as Microsoft to cover its compute spending.
The disclosure arrives as investors and industry observers increasingly scrutinize the economic viability of generative AI labs' business models. The multiyear contracts OpenAI has signed with Nvidia for chips, Oracle for cloud hosting, and CoreWeave for additional compute capacity represent substantial financial obligations that now need to be weighed against the company's actual revenue trajectory.
While the precise underlying figures have yet to be fully confirmed, the revision highlights a broader tension within the AI industry: ambitious growth narratives used to justify unprecedented infrastructure spending versus the more measured reality of near-term revenue generation. Whether this pattern of forward-loaded spending commitments is sustainable at an industry-wide scale remains an open question, and this kind of revision could feed a wider debate over potential overvaluation in the sector.