OpenAI’s revenue is reportedly $20 billion less than previously projected

The Financial Times reports that OpenAI told investors its annualized revenue is approaching $50 billion, about $20 billion below the previously cited figure of nearly $70 billion. The earlier figure was based on investor comparisons with Anthropic, whose revenue calculation includes sales by cloud partners; OpenAI’s does not. The revenue scrutiny comes as OpenAI seeks to justify large investments, after raising $122 billion in a March funding round; its IPO has reportedly been pushed to early 2027.

Analysis

The key risk is not a $20B change in demand, but unreliable comparability and weaker confidence in the metrics used to underwrite AI capital intensity. Anthropic’s inclusion of cloud-partner sales means the prior OpenAI comparison may mix different revenue scopes; it does not establish that either company’s underlying customer demand has moved. Treat both run rates as non-audited indicators, not as a clean competitive ranking.

Near term, expect pressure on private-market marks and greater scrutiny of AI companies’ revenue definitions, cash burn, and compute commitments. Over 1–3 months, the more consequential catalyst is whether OpenAI can demonstrate conversion of usage into collected, recurring revenue while supporting its investment burden. A delay to public-market access increases the importance of private financing conditions, though the reported funding round makes an immediate liquidity conclusion unwarranted. Over 6–18 months, weaker monetization relative to compute needs could shift bargaining power toward cloud providers and constrain smaller model developers; conversely, cloud-partner attribution may obscure rather than reduce total ecosystem sales.

Contrarian read: this is not independently verified evidence of a demand collapse, and the lower figure may chiefly correct a scope mismatch. Public-market exposure is indirect: Microsoft’s economics depend on realized cloud usage and contract terms, not OpenAI’s headline run rate alone. No clean directional public-equity trade follows without exposure and cash-conversion data.

Key Decisions for Investors

• Do not trade the headline as a $20B revenue downgrade until the two companies’ recognition rules and periods are reconciled; require comparable net revenue, collections, and recurring-customer data.
• For private AI marks and new financing, increase the diligence hurdle: track cash burn, committed compute spend, customer concentration, and revenue-to-compute conversion. Reassess if verified disclosures show monetization lagging materially behind capacity commitments.
• Keep Microsoft exposure tied to Azure growth, AI-related utilization, and contract economics rather than OpenAI’s reported run rate. A slowdown in those operating indicators—not this metric dispute alone—would falsify the view that the issue is chiefly comparability.
• No immediate public-market position is warranted on this report alone. Revisit around OpenAI’s next audited financial disclosure, financing update, or IPO timetable change; any deterioration in financing access would raise the risk beyond a valuation-credibility issue.