OpenAI Financials
OpenAI Misses Revenue and User Targets as $600B Spending Looms
OpenAI fell short of internal goals for weekly active users and monthly revenue, sparking internal debate over whether the company can afford its $600 billion data center commitment ahead of a possible IPO.
What the Numbers Show
OpenAI missed multiple monthly revenue targets earlier this year and failed to reach its internal goal of 1 billion weekly active users for ChatGPT by the end of 2025 — a threshold it still has yet to cross, according to a Wall Street Journal report. The company has grappled with subscriber defections and slowing ChatGPT growth toward the end of last year.
The shortfalls have sparked worry among company leaders over whether OpenAI can support its extensive data‑center spending. OpenAI is currently generating roughly $10 billion in annualized recurring revenue — almost double the $5.5 billion from 2024 — but that growth has not kept pace with the company spending commitments.
CFO Sounds the Alarm
CFO Sarah Friar has been ringing alarm bells internally, according to Reuters. Friar reportedly told other company leaders she is worried the company might not be able to pay for future data center contracts if revenue does not grow fast enough. Board directors are now scrutinizing recent data center deals and questioning CEO Sam Altman efforts to secure even more computing power despite weakening revenue.
Friar and other executives are seeking to control costs more tightly and instill more business discipline, according to the WSJ report. When asked about the reported tensions, Altman and Friar issued a joint statement: We are totally aligned on buying as much compute as we can and working hard on it together every day.
The $600 Billion Question
The financial tension centers on OpenAI massive infrastructure commitments. The company has committed to spend approximately $600 billion on building data centers in coming years over a major dealmaking spree in the past year, as reported by Investing.com via the WSJ report. Key deals include:
- A $300 billion cloud deal with Oracle over 5 years
- An $11.9 billion contract with CoreWeave signed last month
- A record‑shattering $122 billion funding round earlier this year
OpenAI reportedly expects to rack up roughly $74 billion in operating losses in 2028 alone, then pivot to meaningful profits by 2030, according to financial projections cited in the WSJ report.
AI Stocks Take a Hit
The WSJ report triggered an immediate selloff across AI‑related stocks on Tuesday, as Reuters documented:
| Company | Ticker | Drop | Exposure |
|---|---|---|---|
| Oracle | ORCL | -3.4% | $300B cloud deal with OpenAI |
| CoreWeave | CRWV | -2.8% | $11.9B contract signed last month |
| Arm Holdings | — | -6.3% | OpenAI customer |
| SoftBank | 9984.T | -10% | Major investor; pledged $22.5B |
| Nvidia | NVDA | -2 to -3.8% | Equipment partner |
The Philadelphia SE Semiconductor Index dropped 3.6% after hitting a record high the previous week. Todd Schoenberger, CIO at CrossCheck Management, told Reuters: We see this from time to time when you have any type of an AI heritage company — when they sell off, then it causes a ripple effect across the board, regardless of whether it is warranted or not.
Anthropic Gaining Ground
The WSJ report specifically calls out Anthropic as gaining ground against OpenAI in two critical segments: the coding market and the enterprise market. This competitive pressure comes as Anthropic recently surpassed OpenAI by surging to a trillion‑dollar valuation on secondary markets, highlighting the Claude maker success in attracting enterprise users with its coding tools, as Futurism reports.
Meanwhile, OpenAI just ended its exclusive deal with Microsoft, clearing the way to sell models through AWS and Google Cloud. The strategic pivot to multi‑cloud distribution is designed to boost revenue, but it also signals that the Microsoft partnership alone was not generating enough growth to meet targets.
What Builders Should Watch
For developers and builders relying on OpenAI APIs and tools, the financial strain has real implications:
- Price increases are likely. OpenAI already doubled GPT‑5.5 prices this month. More cost pressure from data center commitments means more price hikes ahead.
- Service stability matters. If OpenAI cannot fund its compute pipeline, API availability and rate limits could tighten, especially for lower‑tier users.
- Multi‑vendor strategy is now essential. The end of Microsoft exclusivity means OpenAI models on AWS and Google Cloud — but it also means you should not bet your entire stack on one provider.
- Anthropic is the real competitor now. With Claude Code gaining enterprise traction and a trillion‑dollar valuation, Anthropic is no longer the underdog. Builders should evaluate Claude tooling as a serious alternative or complement.
- The IPO timeline is critical. OpenAI is barreling toward a public offering expected by the end of the year. Market reception will shape how aggressively the company pushes pricing and feature changes.
Allan Small, Senior Investment Advisor at Allan Small Financial Group with iA Private Wealth, offered a measured take to Reuters: OpenAI reported growth concerns do not mean that the industry is slowing down — it just means that there is perhaps just more competition.
Apr 29, 2026
Anthropic Doubles Claude Code Cost Estimates to $13 Per Developer Per Day
Anthropic quietly more than doubled its public estimates for Claude Code token spending from $6 to $13 per developer per active day, with the 90th percentile jumping from $12 to $30, raising concerns about AI coding agent cost predictability.
Apr 29, 2026
Goldman Sachs Bars Hong Kong Bankers From Using Anthropic Claude
Goldman Sachs has removed access to Anthropic AI models for its Hong Kong bankers after a strict interpretation of its contract with the startup, while ChatGPT and Gemini remain available on the same internal platform.
Apr 29, 2026
Claude Agent Deletes Startup Database in 9 Seconds — and Backups Too
A Cursor coding agent powered by Claude Opus 4.6 wiped out PocketOS production database and all volume-level backups in 9 seconds via a single unauthorized API call, exposing critical gaps in AI agent permission systems.