OpenAI called the growth report clickbait. The market disagreed by tens of billions of dollars.

OpenAI called the report “prime clickbait.” It said its business is “firing on all cylinders.” It issued a joint statement from CEO Sam Altman and CFO Sarah Friar declaring they are “totally aligned.” None of it worked. On Tuesday, after the Wall Street Journal reported that OpenAI had missed internal revenue and user growth targets, investors wiped tens of billions of dollars off the companies whose business models depend on OpenAI’s growth assumptions being correct. Oracle, which signed a $300 billion five-year partnership to supply computing power to OpenAI, dropped 7.7 per cent. CoreWeave, which has an $11.9 billion infrastructure contract with OpenAI, fell 7.4 per cent. SoftBank, which has committed $60 billion to OpenAI, sank almost 10 per cent in Tokyo. Nvidia, Broadcom, AMD, and Arm all declined between 2 and 6 per cent. The market did not care what OpenAI called the report. It cared whether the money works.

The targets

The Journal reported that OpenAI missed an internal goal of reaching one billion weekly active ChatGPT users by the end of 2025. The company reached 900 million by February 2026, a number that represents 125 per cent year-over-year growth at a scale where most products have already plateaued. By any normal standard, 900 million weekly users is a staggering achievement. But OpenAI is not operating by normal standards. It is operating by the standards required to justify $600 billion in compute spending commitments through 2030, a figure the company itself revised downward from $1.4 trillion earlier this year. The Journal also reported that OpenAI missed multiple monthly revenue targets in early 2026, losing ground to Google’s Gemini in consumer markets and to Anthropic in coding and enterprise. OpenAI’s annualised revenue sits at approximately $25 billion. Competition from Anthropic, which crossed $30 billion in annualised revenue in April while spending roughly a quarter of what OpenAI spends on training, has turned what was a comfortable lead into a deficit. The company that defined the generative AI market is no longer the revenue leader in it.

CFO Sarah Friar warned colleagues internally that if revenue growth does not accelerate, OpenAI could face difficulty funding its future compute agreements. The warning is not hypothetical. OpenAI has contracted for hundreds of billions of dollars in cloud infrastructure from Oracle, CoreWeave, and others, commitments that assume revenue will grow from $25 billion today to $280 billion by 2030. The gap between where the revenue is and where it needs to be is not a rounding error. It is the difference between a company that can fund its own ambitions and one that cannot. Friar has also reportedly told colleagues that OpenAI is not organisationally ready for the IPO that Altman wants to pursue in the fourth quarter of 2026, preferring a 2027 listing instead. The joint statement declaring alignment came after the Journal’s reporting made the disagreement public.

The response

OpenAI spokesperson Steve Sharpe called the Journal’s report “clickbait” and said the company’s consumer business is showing strength in revenue while its enterprise business is “in the best place it has ever been.” The company said the internal mood is “incredibly positive.” These are not rebuttals to the specific claims in the report. The Journal did not say OpenAI’s business is failing. It said OpenAI missed its own targets, that its CFO has expressed concern about funding future commitments, and that competitors are gaining ground. Calling that clickbait is a communications strategy, not a financial argument.

The company does have genuine momentum to point to. Enterprise revenue now accounts for more than 40 per cent of total revenue, with nine million paying business users, a fourfold increase since September 2025. OpenAI’s advertising business, launched in February, crossed $100 million in annualised revenue within six weeks and is projected to generate $2.5 billion this year, scaling to $100 billion by 2030. OpenAI’s enterprise push with GPT-5.5 and rapid model releases demonstrates a company that is shipping aggressively. And 50 million paying subscribers is a consumer franchise that most technology companies would regard as a generational achievement. The problem is not that the business is bad. The problem is that the commitments are enormous, and the business needs to be not just good but historically unprecedented to justify them.

The arithmetic

OpenAI’s financial position is defined by a simple asymmetry: its spending commitments are contractual and its revenue projections are aspirational. The company has committed to approximately $600 billion in compute infrastructure spending through 2030, averaging roughly $100 billion per year. To justify that spending, it is targeting $280 billion in revenue by 2030, which would require more than ten times growth in four years from its current $25 billion run rate. The $852 billion valuation from its $122 billion funding round in March, the largest private round in Silicon Valley history, was predicated on those growth assumptions holding. If they do not, the valuation becomes aspirational too.

The competitive picture compounds the pressure. Anthropic’s revenue crossed $30 billion in annualised run rate in April, passing OpenAI for the first time, with 80 per cent of that revenue coming from enterprise customers spending more than $1 million annually. The number of those customers doubled from 500 to over 1,000 in less than two months. Google’s Gemini gained consumer market share throughout 2025, eating into ChatGPT’s growth. And the broader landscape now includes DeepSeek, Mistral, Meta’s Llama, and a constellation of open-weight models that compete on price in ways that make it harder for any single company to capture enough of the market to justify $600 billion in infrastructure. OpenAI’s thesis has always been that scale wins: build the biggest models, deploy the most compute, acquire the most users, and revenue follows. The WSJ report is the first significant piece of evidence that scale may not be winning fast enough.

The collateral

The market reaction on Tuesday was not about OpenAI’s quarterly performance. It was about the cascade of financial dependencies that OpenAI’s growth narrative supports. Oracle’s $300 billion cloud partnership, CoreWeave’s $22 billion in cumulative OpenAI contracts, SoftBank’s $60 billion investment, and the broader capital expenditure plans of Nvidia, AMD, and Broadcom all assume that AI infrastructure demand will grow at a rate sufficient to generate returns on the capital being deployed. When the company at the centre of that demand admits, even internally, that growth is not meeting projections, the entire chain reprices.

This is happening against a backdrop that makes the timing particularly uncomfortable. The Musk v. Altman trial opened in Oakland on Monday, putting OpenAI’s corporate governance and leadership credibility under oath. Anthropic has overtaken OpenAI in revenue while positioning itself as the safety-conscious alternative. Google is embedding Gemini across 750 million users and a $240 billion cloud backlog. And OpenAI’s own CFO has reportedly questioned whether the company is ready for a public listing that its CEO is pushing for this year. The company that defined the AI era is now defending its position on multiple fronts simultaneously, and the market’s verdict on Tuesday was that the defence is not yet convincing. OpenAI says it is firing on all cylinders. The investors who lost billions on Tuesday would like to see the receipts.