A $20 billion discrepancy in OpenAI’s reported revenue figures has put US technology stocks under pressure. Nvidia, Oracle and other beneficiaries of the AI boom suffered significant losses. However, new information suggests that the apparent shortfall may be partly explained by different calculation methods. The episode highlights just how sensitive financial markets have become to the growth expectations surrounding a handful of major AI companies.
A Report Sends the Nasdaq Lower
US technology stocks came under selling pressure on Thursday. The Nasdaq Composite fell 1.25%, while the broader S&P 500 declined 0.47%. Companies whose growth prospects are closely linked to the expansion of artificial intelligence were particularly affected.
Nvidia shares lost 2.9%, while Oracle fell approximately 5.5%. Broadcom, Micron and other semiconductor companies also recorded substantial declines.
The immediate trigger was a Financial Times report on OpenAI’s revenue development. According to the report, the developer of ChatGPT had reached an annualized revenue run rate of nearly $50 billion by the end of September. Earlier media reports citing information from investor circles had suggested a figure approaching $70 billion.
The difference of roughly $20 billion caused considerable uncertainty. OpenAI is widely regarded as one of the most important drivers of the global AI boom. The company not only develops leading language models but is also a major customer for computing power, cloud services and specialized hardware.
If such a significant customer generates less revenue than expected, the growth assumptions underpinning its numerous business partners may also come under pressure.
However, the OpenAI report was not the only factor weighing on financial markets. Rising oil prices, inflation concerns and uncertainty over the future direction of interest rates added to investors’ nervousness. The Nasdaq’s decline therefore cannot be attributed exclusively to the newly reported revenue figures.
$50 Billion or $70 Billion?
The crucial question is what the two figures actually represent.
Neither the $50 billion nor the $70 billion refers to annual revenue already generated. Instead, both figures concern annualized revenue run rate.
This metric extrapolates revenue generated over a shorter period to a full year. For example, if a company earns $5 billion in one month, multiplying that figure by twelve produces an annualized revenue run rate of $60 billion.
Fast-growing technology companies frequently use this measure to illustrate their current business momentum. For a company whose revenue is increasing rapidly from month to month, the annualized figure can be significantly higher than the revenue actually generated over the preceding twelve months.
The metric therefore provides valuable information, but it is no substitute for a comprehensive income statement.
According to the Financial Times, the discrepancy between the previously circulated $70 billion figure and the newly reported $50 billion figure also arose from attempts to compare the business performance of OpenAI and its competitor Anthropic directly.
The two companies reportedly treat revenue generated through cloud partners differently. While Anthropic includes revenue distributed through providers such as Amazon Web Services and Google Cloud in its corresponding figures, OpenAI does not account for such partner revenue in the same way.
Attempts by investors to adjust the numbers for comparison are said to have produced higher revenue estimates for OpenAI.
This does not mean that $20 billion is actually missing from the company’s accounts. Rather, different definitions of the same underlying business activities can result in substantially different financial metrics.
New Information Calms the Markets
Another report emerged on Friday, putting the initial concerns into perspective.
Bloomberg reported, citing people familiar with the financial figures, that OpenAI still expected to reach an annualized revenue run rate of at least $70 billion by the end of 2026. At the end of September, the corresponding figure had reportedly stood at approximately $50 billion.
Contrary to what the initial market reaction suggested, these two figures are not necessarily contradictory. One describes the position at the end of September, while the other represents the revenue level expected by year-end.
Growth is reportedly being driven particularly by the enterprise customer business. Reaching the target would require OpenAI to increase its current monthly revenue pace substantially during the final quarter.
Whether the company will achieve this goal remains unproven. The figures are based on information from investor circles rather than a published, independently audited annual financial statement.
Nevertheless, the new reports provided some relief. Nasdaq futures temporarily indicated a significant recovery during Friday’s pre-market trading.
What remains remarkable is that a debate over the calculation of a revenue metric was enough to trigger substantial share-price movements across numerous publicly traded companies.
Why OpenAI Moves the Entire AI Sector
OpenAI itself is not publicly listed. Nevertheless, its financial prospects increasingly influence the valuations of numerous publicly traded technology companies.
Nvidia supplies the high-performance processors required to train and operate large AI models. Oracle is investing heavily in cloud and data center infrastructure. Microsoft has closely integrated artificial intelligence into its cloud services and software products and maintains an extensive partnership with OpenAI.
Other beneficiaries include semiconductor manufacturers, networking equipment suppliers, energy providers and companies involved in planning, financing and operating data centers.
Their economic prospects do not depend exclusively on OpenAI. Demand for AI infrastructure is also driven by other developers, major technology corporations and numerous enterprise customers. However, OpenAI remains one of the most visible companies in the market and plays a central role in several major infrastructure projects.
For investors, this creates a particular situation: They cannot directly purchase shares in the privately held company, but they can invest in suppliers and business partners whose expected profits depend on growing demand for artificial intelligence.
The share prices of these companies therefore reflect not only current earnings but also assumptions about the future growth of the entire industry.
When those expectations are called into question, reassessments can quickly spread across multiple companies simultaneously.
The Investments Still Have to Pay Off
The expansion of artificial intelligence represents one of the most capital-intensive investment programs in the history of the technology industry.
Large language models require enormous computing capacity. Operating them involves expenditure on specialized processors, servers, networking equipment, electricity, cooling systems and the construction of new data centers.
Research, personnel and the development of new generations of AI models add further costs.
Traditional software companies can often benefit from relatively low additional costs when serving more users. By contrast, providing advanced AI services initially involves substantial ongoing consumption of computing resources.
Economic viability therefore depends not only on how many people use ChatGPT or comparable applications. Equally important is how much revenue this usage generates and what it costs to provide the necessary computing power.
The calculation becomes particularly challenging when companies must finance large portions of their future infrastructure today, even though the anticipated revenue may not materialize until several years later.
Morgan Stanley has estimated that global AI infrastructure expansion could require approximately $2.9 trillion in investment over a multi-year period. According to the bank, this could create a financing gap of around $1.5 trillion that would need to be covered by external capital.
This figure is not a forecast of losses. It represents the estimated financing requirement beyond the funds that major technology companies may be able to provide themselves.
The more heavily data center expansion relies on debt financing, the more important interest rates, financing conditions and the reliability of future revenues become.
Investors are therefore no longer focusing exclusively on the technological potential of artificial intelligence. They are increasingly asking when, and to what extent, these enormous investments will generate economic returns.
A Valuation of $1.4 Trillion
The debate over OpenAI’s revenue figures comes at a time when the company is reportedly seeking to raise substantial additional capital.
According to Bloomberg, OpenAI is holding discussions about a funding round of at least $30 billion, with a potential pre-money valuation of approximately $1.4 trillion.
Such a valuation would reflect exceptionally high expectations for the company’s future market position and earnings potential.
However, it says relatively little about current profitability. As with other rapidly growing technology companies, investors are valuing not only existing revenue but also the possibility of capturing a significant share of an emerging multibillion-dollar market.
A realistic assessment therefore requires consideration of spending on computing power, research and infrastructure alongside revenue growth. Equally relevant is how much of that revenue can ultimately be converted into sustainable profits after costs have been deducted.
The reported increase in OpenAI’s annualized revenue run rate to approximately $50 billion within a relatively short period points to extraordinarily strong demand. Yet it does not answer the question of when the company’s substantial investments will translate into sustainably positive financial results.
Uncertainty is compounded by the fact that privately held companies are subject to significantly less extensive financial disclosure requirements than publicly traded corporations.
Until OpenAI publishes comprehensive audited financial statements, outside observers will remain dependent on company disclosures, information from investors and journalistic investigations when assessing important financial metrics.
The AI Boom Is Being Measured by Revenue
The recent share-price movements do not prove that demand for artificial intelligence is collapsing. On the contrary, even the newly reported annualized revenue run rate of approximately $50 billion represents extraordinary business growth.
Nor do different calculation methods automatically imply that earlier figures were deliberately misleading.
The episode does, however, illustrate how sensitively financial markets react to information that appears inconsistent with exceptionally high growth expectations.
The economic significance of artificial intelligence is no longer judged solely by technological capabilities. Increasingly, the decisive factors are measurable revenue, the profitability of business models and whether growing demand can sustainably justify billions of dollars in infrastructure investment.
This creates a tension for stock markets: The potential profits from artificial intelligence are enormous, but a substantial portion of those expectations is already reflected in the valuations of many technology companies.
Thursday’s correction and the subsequent recovery demonstrate how quickly market sentiment can shift, even though the underlying technology has barely changed within the space of a single day.
The real risk is not simply that OpenAI is currently generating less revenue than one previously circulated estimate suggested. It is that financial markets are increasingly measuring the development of an entire industry against growth figures whose calculation and comparability still require careful explanation.
SK
