What the Wall Street Journal report on OpenAI financials reveals and why it triggered a sell-off in AI-related stocks
On April 27, 2026, the Wall Street Journal published a report stating that OpenAI had missed several internal targets for revenue and user growth in recent months. According to the publication, CFO Sarah Friar reportedly told colleagues she was concerned about the company ability to fund future compute contracts if revenue growth does not accelerate.
The report cites people familiar with the matter and indicates that OpenAI missed multiple monthly sales targets in 2026, after losing ground to Anthropic in the coding and enterprise markets. According to the same WSJ article, ChatGPT growth slowed toward the end of 2025 and the company fell short of an internal goal to reach 1 billion weekly active users by year-end.
The key point. The report is not about a single quarterly loss but about the relationship between actual revenue growth and the future spending commitments already signed. It is this relationship that the market is now repricing.
To frame the scope of the issue, it is useful to look at the publicly available numbers, as reported by WSJ, Reuters, Bloomberg and the official communications of the involved suppliers.
According to WSJ, OpenAI has committed to spend approximately $600 billion on building data centers in the coming years. Based on financial projections cited in the WSJ report, the company expects roughly $74 billion in operating losses in 2028 alone, before pivoting to meaningful profits by 2030.
The $122 billion funding round completed in 2026 represents according to Reuters a sector record, but it remains a fraction of the compute commitments already signed.
The $600 billion in compute commitments is composed of multi-year agreements with three primary suppliers. The figures below are those publicly reported by the parties or cited by WSJ and other financial media.
| Supplier | Amount | Duration | Announcement |
|---|---|---|---|
| Microsoft Azure | $250B | about 6 years (2025 to 2030) | October 2025 |
| Oracle | $300B | 5 years (2027 to 2031) | September 2025 |
| Amazon Web Services | $38B | Not disclosed | November 2025 |
| CoreWeave | $11.9B | Multi-year | March 2026 |
The Oracle deal, $60 billion per year for five years starting in 2027, was confirmed by CEO Safra Catz on Oracle earnings calls and is part of the Stargate data center buildout. Oracle remaining performance obligations (RPO) grew from $455 billion in Q1 FY2026 to $523 billion in Q3 FY2026 according to data published by the company.
The Microsoft commitment of $250 billion was announced in October 2025 as an incremental commitment on Azure services. The AWS deal of $38 billion, reported in November 2025, ended Microsoft exclusivity as cloud provider.
For context, according to TechCrunch, building 1 GW of data center capacity costs approximately $50 to $60 billion. The total $600 billion in commitments implies a compute capacity historically unprecedented for a single company.
On the day the report was published, April 28, 2026, stocks tied to OpenAI supply chain registered significant declines. The figures below are reported by CNBC and NBC News.
SoftBank closed down 10% in Tokyo. The Japanese conglomerate has committed $60 billion in OpenAI investments.
Oracle opened down approximately 5.5%. CoreWeave dropped 5.4 to 7%. The company has signed $11.9 billion in contracts with OpenAI.
Nvidia became the worst performer in the Magnificent 7, down more than 1%. The original Nvidia OpenAI deal of up to $100 billion from September 2025 was reduced to $30 billion according to the Financial Times.
S&P 500 closed down 0.49% at 7,138.80. Nasdaq lost 0.9% at 24,663.80. AMD and Broadcom dropped about 4 and 3%.
As an analysis from Sherwood News published after the report notes, the dynamic observed in these stocks reflects their function as public indicators of OpenAI demand. When doubts emerge about the company growth trajectory, the entire compute supply chain is repriced together.
OpenAI publicly pushed back on the report. According to CNBC, the company stated: "This is ridiculous. We are totally aligned on buying as much compute as we can and working hard on it together every day". NBC News also reports that head of business and financial communications Steve Sharpe called the WSJ report "clickbait" and said the company was "on an extremely steep growth curve across consumer, enterprise and developers".
According to Fortune, Gene Munster of Deepwater Asset Management argued that OpenAI business is growing rapidly and could potentially double year over year. Munster pointed to Codex, the coding tool based on GPT 5.5, as an underestimated revenue line. Deepwater holds a position in OpenAI through private markets.
Oracle defended OpenAI growth trajectory with an official statement cited by CNBC: the company said it was "incredibly excited" about the partnership and focused on delivering the capacity needed to support demand.
John Belton, portfolio manager at Gabelli Funds, told CNBC that OpenAI growth had slowed in late 2025 and early 2026, losing market share to Anthropic and Gemini. However, Belton specified that this does not necessarily imply an issue with the pace of spending of the entire sector, but rather a confirmation about OpenAI recent market share trends.
Luke Rahbari of Equity Armor Investments, also cited by CNBC, urged caution on revenue projections in the AI sector, given that forecasts remain imprecise in a rapidly evolving industry.
Worth noting. NBC News specified that it had not independently verified the WSJ reporting. This is an important detail for any reader: the debate over the numbers is still ongoing.
The WSJ report places the loss of market share to Anthropic as one of the factors behind the missed targets. According to the WSJ and Bloomberg, Anthropic gained ground in the coding and enterprise segments.
According to Sherwood News, OpenAI CEO Sam Altman reportedly called for a "code red" to improve ChatGPT in late 2025. The company then doubled down on Codex and on greater compute availability as growth levers for 2026, especially in the enterprise segment.
| Event | Date | Notes |
|---|---|---|
| Oracle deal announcement | September 2025 | $300B over 5 years |
| Microsoft Azure commitment | October 2025 | $250B incremental |
| AWS deal | November 2025 | $38B, end of Microsoft exclusivity |
| Internal "code red" | Late 2025 | Reported by Sherwood News |
| CoreWeave contract | March 2026 | $11.9B |
| WSJ report | April 27, 2026 | Miss on revenue and user targets |
| Musk vs Altman trial | April 27, 2026 | $130B in damages requested |
On the legal front, the Musk vs Altman trial opened in California on April 27, 2026. According to Fortune, Musk is seeking the unwinding of OpenAI for-profit conversion and over $130 billion in damages. The case adds further uncertainty in the period leading up to the company expected IPO, valued at approximately $1 trillion according to Reuters.
The WSJ report has surfaced a tension that was already latent in the markets: the gap between OpenAI current revenue and the spending commitments signed for the coming years. The positions of the parties are far apart. The WSJ reports internal concerns, OpenAI denies and calls the report clickbait, some analysts call the market reaction excessive, others see confirmation of a trend of share loss to competitors.
The publicly available numbers are the following: $600 billion in future compute commitments, $250 billion with Azure, $300 billion with Oracle, $38 billion with AWS, $11.9 billion with CoreWeave, expected IPO valuation around $1 trillion, expected operating losses of $74 billion in 2028. The question the market is trying to price is whether OpenAI future revenue will be sufficient to cover these commitments.
For those following the AI sector as investors or analysts, this episode shows how the relationship between revenue and capex has become the main thermometer for the sector. A change in expectations about OpenAI growth moves not only the stock of the funder (SoftBank), but also cloud suppliers (Oracle, CoreWeave), chip makers (Nvidia, AMD, Broadcom) and, in aggregate, the technology indexes.