Volatility struck U.S. stock markets last week following a revision in OpenAI’s reported annualized revenue, highlighting the challenges investors face in interpreting financial metrics used by rapidly growing artificial intelligence companies. OpenAI disclosed that its annualized net revenue stands at approximately $50 billion, about $20 billion less than the $70 billion figure widely circulated in late September. This adjustment has drawn attention to the use of annualized revenue—a forward-looking estimate based on recent performance—as a key but sometimes ambiguous metric among AI start-ups.

OpenAI and its rival Anthropic, both private entities competing to develop advanced AI models, have become focal points for investor sentiment regarding the AI sector’s growth prospects. The companies’ financial data are viewed as barometers for the broader AI boom, which has influenced stock prices of related firms including Nvidia, Microsoft, Oracle, and others linked to the AI supply chain.

The $70 billion figure initially reported was described by OpenAI as a “net” revenue metric, reflecting the company’s share of proceeds after accounting for partnerships such as those with Microsoft. In contrast, Anthropic calculates its annualized revenue on a gross basis, including the full value of sales made through cloud partners. This difference in accounting practices has contributed to confusion and difficulties in making direct comparisons between the two companies.

OpenAI clarified that its internal accounting consistently uses the net revenue figure and that even on a gross basis, the earlier $70 billion estimate would be inaccurate. The company also recently highlighted a strong quarter with 70% growth in its annualized run rate revenue, and revised upward its estimated revenue for 2026 from $30 billion to $35 billion, partly driven by the release of its GPT-5.6 model in July.

The uncertainty surrounding these annualized figures has underscored the need for more standardized financial disclosures. While current investors are better positioned to seek clarification about the metrics and methodologies used, outsiders and market observers face challenges interpreting the information, leading to fluctuations in stock prices for companies tied to AI’s trajectory. For example, the tech-heavy Nasdaq 100 index fell 1.4% last Thursday amid the revenue uncertainty but partially recovered the following day.

Market experts and analysts have emphasized that these annualized revenue numbers, which extrapolate near-term results to a full year, are inherently imperfect and can be influenced by complex revenue models involving token usage, subscriptions, and licensing agreements, complicating revenue recognition. Despite these limitations, such figures remain vital for venture capitalists and private investors looking to gauge start-up momentum, where traditional profitability metrics might be less relevant.

Both OpenAI and Anthropic face pressure to pursue initial public offerings, which would require regulatory disclosures providing more detailed and standardized financial reporting. Anthropic is anticipated to make this move later in 2024, while OpenAI has postponed its public listing to next year. Greater transparency could reduce investor uncertainty, enabling more accurate pricing of AI-related securities and diminishing market volatility tied to headline revenue estimates.