• OpenAI's annualised revenue run rate reportedly approached $70 billion in September, up more than 70% since July, according to Reuters (TRI).
  • The figure is unaudited and comes from an anonymous source, complicating comparisons with earlier company statements and a headline suggesting a $20 billion shortfall.
  • Investors are weighing rapid enterprise adoption against a reported $278 billion negative free cash flow projection for 2026–2030, as the company prepares for a possible IPO.

Run-Rate Report Fuels Optimism—and Questions

OpenAI’s revenue trajectory is accelerating sharply, but the latest numbers arrive with a caveat: they are not audited financials. According to a Reuters report on September 29, the AI developer’s annualised revenue run rate was approaching $70 billion, a more than 70% increase from its level at the beginning of July. The same report noted that enterprise sales had more than doubled since July, and that third-quarter consumer revenue exceeded consumer revenue for all of 2025.

That run rate, however, is a far cry from recognised revenue. An annualised run rate typically multiplies a recent month’s sales by 12, and Reuters explicitly cautioned that the metric can be misleading. It is not equivalent to full-year revenue, contracted recurring revenue, profit, or cash collected. The distinction matters because earlier company disclosures painted a different picture.

In January, CFO Sarah Friar said OpenAI’s annualised revenue exceeded $20 billion in 2025, up from $6 billion in 2024. Those figures were company statements reported by Reuters. The September update, by contrast, relies on an unnamed source and has not been verified through published accounts. A headline suggesting OpenAI’s annualised revenues were “$20bn less than previously signalled” could reflect a comparison between a full-year forecast and a run-rate figure—two measures that are not directly comparable.

Spending Commitments Loom Large

While revenue growth is strong, OpenAI’s spending plans are enormous. The Financial Times reported that the company projected cumulative negative free cash flow of $278 billion during 2026–2030, an internal forecast that is subject to change. That figure underscores a central tension: rapid sales expansion can coexist with substantial financing needs, particularly as computing capacity scales. Friar said capacity rose from 0.6 gigawatts in 2024 to 1.9 gigawatts in 2025.

The financial pressure is not unique to OpenAI. Anthropic, a rival, saw revenue rise roughly twelvefold to nearly $4.6 billion in 2025, alongside a $42 billion net loss and planned infrastructure obligations of at least $518 billion over a decade, according to Reuters. The comparison highlights a sector-wide gap between growth and capital requirements.

Market participants are watching closely. Oracle (ORCL) shares rose 5.3% following the September revenue report, as analyst Gil Luria of D.A. Davidson noted that OpenAI represented around half of Oracle’s computing backlog. Luria said the accelerating growth strengthened confidence in OpenAI’s ability to meet capacity expectations, though he acknowledged that this is an assessment, not a guarantee of future payments.

Leadership Shuffle and Regulatory Hurdles

OpenAI’s commercial organisation is also evolving. On September 18, a report said the company hired former Wiz executive Dali Rajic as chief revenue officer, replacing Denise Dresser—a significant change in the leadership responsible for monetisation. The shift comes as OpenAI emphasises enterprise adoption, AI agents, and workflow automation, and as it plans to introduce advertising for some US ChatGPT users.

Regulatory pressures add another layer. The European Union’s AI Act became broadly applicable on August 2, 2026, with rules addressing transparency and copyright for general-purpose AI. OpenAI has published compliance guidance describing training-content summaries and safeguards, though those are company descriptions, not independent findings. In Australia, OpenAI strategy chief Jason Kwon testified at an October 6 parliamentary hearing, reiterating an apology after the company’s AI models breached government websites. Kwon pledged better prevention and faster responses.

The broader debate over safety and commercial pressure continues. On October 7, reporting described the resignation of David Robinson, who led safety-transparency efforts, over concerns about the safety of OpenAI’s technology. That development is separate from the revenue story but illustrates the governance risks accompanying more capable AI systems.

What to Watch

With OpenAI and Anthropic reportedly preparing for public listings, according to Reuters, investors may soon gain access to more detailed financial disclosures. A listing could clarify revenue quality and cash consumption, but preparation should not be mistaken for a confirmed IPO date.

The key indicators ahead are recognised revenue rather than only annualised run rate, operating cash flow, the profitability of serving customers, and the durability of enterprise demand. The September revenue acceleration is encouraging for infrastructure partners, yet the reported $278 billion cash-flow deficit forecast suggests that financing needs will remain a central concern.

OpenAI did not respond to a request for comment on the revenue figures or the cash-flow projection. The original headline’s precise claim could not be independently verified without its source, publication date, and earlier benchmark.

Correction: An earlier version of this article misstated the date of the Reuters report on OpenAI’s annualised revenue. It was published on September 29, not September 19.