OpenAI Trims 2026 Revenue Expectation to $50bn, $20bn Below Earlier Signal
OpenAI has told investors it expects 2026 revenue of $50bn (£37bn), about $20bn less than the $70bn it had signalled in September.
The revised figure was based on sales up to the end of September 2026. The earlier $70bn indication had been provided in information to investors in September 2026. That prior estimate was derived by applying 70% revenue growth to an initial $40bn estimate, according to Yahoo Finance.
The update emerged while OpenAI is in early-stage talks to raise $30bn in a funding round valuing the business at about $1.4tn, according to The Guardian. The talks are not concluded. The valuation discussion and the revenue guidance are proceeding in parallel.
Markets reacted on Thursday 8 October 2026. The tech-led Nasdaq closed down 1.4% after news of the gap. Oracle shares fell 5.5%. Micron declined 4.8%. Nvidia fell 2.9%.
Part of the comparison set turns on accounting. Anthropic includes revenue from sales via cloud partners such as Amazon's AWS and Google Cloud, while OpenAI does not. The distinction affects how headline revenue figures compare across labs, even when underlying demand may differ from reported totals.
The broader context here is calibration rather than direction. A $20bn revision to a forward-year estimate changes the denominator against which a proposed valuation is judged. For investors weighing an early-stage raise, the question is not only the absolute growth rate but the stability of the forecasting method. A projection built by applying a fixed growth percentage to a base estimate behaves differently from one anchored in nine months of booked sales. Expert readers will recognize the difference. The former rewards assumptions. The latter tests them.
Looking at what this means for the next stage, attention is likely to focus on two variables. First, how investors normalize for channel differences when benchmarking OpenAI against peers that book cloud-partner sales differently. Direct comparison without adjustment risks misreading performance. Second, whether the September and October figures are treated as successive refinements of the same forecast or as distinct methodologies. That choice will shape due diligence.
In my view, the market response is the element to watch. Listed infrastructure names moved together on the news. That linkage suggests AI-lab guidance now functions as a near-term pricing input for public equities, not only as a private-market signal. If that holds, precision in investor communications carries wider consequences. Small shifts in language around estimates can move large pools of capital. The incentive, for labs and for investors, is toward clearer disclosure of bases, cut-off dates and inclusions.
