OpenAI sees 2026 run rate revenue reaching or topping USD 70bln, according to Bloomberg sources

Sourced revenue figures for private AI labs have become a recurring fixture of the tape, typically surfacing around funding negotiations or secondary share sales, where the incentive to circulate an aggressive run rate is strongest. The distinction that matters is between run rate and recognised revenue: an annualised snapshot of a single month's billings flatters the trajectory relative to audited figures, and in past episodes of this kind the gap between the two has been material once cohort churn and promotional pricing wash through. For listed markets the transmission is indirect and runs through the AI complex's revenue visibility, the hyperscalers carrying OpenAI compute commitments, the cloud and infrastructure names with exposure to its capacity bookings, and the chip suppliers whose order books implicitly assume this growth path. Projections of this sort have historically served to validate or stress the circular financing logic underpinning the sector, where lab revenue underwrites cloud capex which in turn underwrites lab compute. Worth noting is whether the figure is corroborated by named financials in any subsequent funding documentation rather than remaining a sources-based leak. The read-through to peers is directional only; the number itself is unverifiable until the company puts accounts on the record.