Rapid expansion is colliding with the extraordinary price of technological leadership.
SAN FRANCISCO, UNITED STATES — June 2026. OpenAI closed 2025 with approximately $13 billion in revenue, more than tripling the $3.7 billion generated a year earlier. However, the company also reported a net loss of nearly $39 billion, exposing the vast financial cost of developing and operating advanced artificial intelligence systems at global scale.
Revenue growth was driven by paid ChatGPT subscriptions, corporate agreements and the increasing use of OpenAI models through application programming interfaces. These tools allow businesses and developers to integrate artificial intelligence into their own products and services, expanding the company’s commercial reach beyond individual users.
OpenAI’s expenses rose even faster than its income as the company continued investing heavily in computing infrastructure, specialised personnel, model development and international expansion. Total costs increased substantially during the year, reflecting the scale of the resources required to remain competitive in the global artificial intelligence race.
Research and development represented the company’s largest expense, reaching approximately $19.2 billion in 2025, compared with $7.8 billion the previous year. These costs included model training, cloud infrastructure, advanced computing capacity and the recruitment of highly specialised researchers and engineers.
OpenAI reportedly paid Microsoft approximately $10.6 billion for services classified within research and development, including computing resources provided through the Azure cloud platform. The relationship remains central to OpenAI’s operations because training and running frontier models requires access to enormous amounts of processing power.
The cost of operating ChatGPT and other products also increased sharply as the volume of user requests expanded. Cost of revenue reached around $7.5 billion, while sales and marketing expenditure approached $5.7 billion as the company sought to attract more corporate clients and strengthen its position in international markets.
General and administrative expenses also rose as OpenAI expanded its legal, security, compliance and organisational structures. The company has had to reinforce these areas while facing greater scrutiny from regulators, investors and governments over the social and economic impact of artificial intelligence.
The final net loss was considerably larger than the operating deficit because the financial statements included major non-cash accounting adjustments. OpenAI recorded an operating loss of approximately $20.9 billion in 2025, compared with about $8.8 billion in 2024.
The final result moved closer to $39 billion after charges related to investor warrants, convertible interests, stock-based compensation and other valuation adjustments were included. These entries did not necessarily represent cash leaving the company during the same year, but they significantly affected the reported financial result.
Even after separating those accounting effects, the figures confirm that OpenAI continues to consume enormous amounts of capital. The company remains dependent on external financing and strategic partnerships to sustain its research, infrastructure and commercial expansion.
Despite the losses, OpenAI’s business momentum accelerated throughout 2025. The company expanded beyond individual ChatGPT users by promoting tools for software development, education, healthcare, finance, research, customer service and public-sector operations.
Its enterprise business has become increasingly important because institutional clients can generate larger and more predictable contracts than individual subscriptions. OpenAI is therefore seeking to transform widespread public adoption into recurring corporate revenue.
The company entered 2026 with substantial financial resources supported by strategic partnerships and successive investment rounds. These funds give OpenAI the capacity to continue investing despite its negative results, but investors will increasingly expect evidence that its business model can eventually become sustainable.
The figures expose the main financial challenge facing the generative artificial intelligence industry. Demand is expanding rapidly, but the infrastructure required to satisfy that demand remains exceptionally expensive.
OpenAI must continue investing in data centres, advanced chips, electricity, cloud services, cybersecurity and specialised talent while competing with Google, Anthropic, Meta and xAI. Reducing research spending could improve short-term financial performance, but it could also weaken the technological advantage that attracts users, customers and investors.
Raising prices could improve margins, although it might encourage some customers to move toward less expensive alternatives. OpenAI’s path to profitability will therefore depend on making its models more efficient, diversifying its sources of income and converting extraordinary technological growth into a sustainable business structure.
Artificial intelligence leadership carries an unprecedented financial price.