OpenAI in 2025: losses soared to $38.5 billion — what is behind the record losses?
OpenAI's financial report for 2025 paints a frightening picture: the company's net loss was approximately $38.5 billion. This is nearly eight times greater than in 2024, when losses were at $5.09 billion. The figures I analyzed speak for themselves — this is a classic case of hypertrophied cost growth against a backdrop of rapid, but insufficient, revenue increases.
In 2024, OpenAI's revenue was $3.7 billion with total expenses of $12.48 billion. The operating loss then reached $8.78 billion. In 2025, revenue jumped to $13.07 billion, but expenses soared to $34 billion. The operating deficit grew to $20.92 billion. The key driver was colossal spending: $19.18 billion on research and development, as well as $7.5 billion on cost of revenue and $5.73 billion on sales and marketing.
Restructuring and Hidden Losses
OpenAI's transition from a non-profit structure to a commercial one resulted in a one-time loss of $41.55 billion due to the revaluation of convertible instruments and warrants. Excluding these items, the net loss would have been $60.35 billion. However, the company managed to "soften" the figure to $38.53 billion by excluding $17.87 billion and an additional $3.95 billion in items related to non-controlling interests.
Payments to Microsoft: The Cost of Infrastructure
The relationship with Microsoft deserves special attention. In 2025, OpenAI paid the tech giant $10.59 billion under "research and development" (likely the cost of training models), $6.047 billion on cost of revenue, and $527 million on sales and marketing. In total — $17.2 billion. At the end of the year, liabilities to Microsoft reached $3.64 billion.
These figures lay bare the fundamental economics of the entire AI industry: building cutting-edge models requires astronomical investments in computing power, which currently far outpace revenue. The main source of losses is not salaries or research, but the cost of servicing billions of daily queries on expensive chips. OpenAI has already contracted approximately $600 billion in future data center expenses.
Analyst's Comment: OpenAI's financial performance is not just the story of one company. It is a stress test for the entire AI industry. The scale of investment required to maintain leadership calls into question profitability in the foreseeable future. The market may be overestimating the speed of monetization for generative models, and investors should prepare for a long period of "scorched earth" in the balance sheets of leading players.