Reports citing documents disclosed to shareholders reveal that OpenAI, the developer of ChatGPT, recorded a cash burn of $3.7 billion in the first quarter of 2026. This figure represents more than half of its quarterly revenue of $5.7 billion for the period.
This disclosure follows recent reports that the company's 2025 expenditures reached approximately $34 billion as it prepares for a potential public listing in the U.S. stock market. These latest operational figures raise questions among investors regarding the financial sustainability and fundamental growth trajectory of the world's highest-valued AI application company.
From a positive perspective, the robust revenue growth underscores a genuine and explosive global demand for AI applications. However, it also indicates that leading AI model companies have entered a capital-intensive arms race phase. This stage is characterized by the coexistence of strong revenue growth, high capital consumption, and significant depreciation pressures, often described as a period of "revenue surge alongside a loss black hole."
OpenAI Continues Aggressive Spending Pre-IPO
The shareholder documents indicate that the continued surge in spending reflects massive ongoing investments in AI computing infrastructure. These investments are necessary to meet the soaring demand for AI training and inference computing power, covering areas such as model training, AI data center construction, talent acquisition, and market share expansion.
The financial data also shows that these substantial expenditures widened the net loss from $5 billion in 2024 to approximately $39 billion in 2025. Despite the high cost base associated with AI computing, OpenAI achieved strong revenue growth of around $13 billion during the same period, representing exponential growth compared to previous years. This growth was primarily driven by an increase in subscription-paying users for its premium ChatGPT and large AI model services.
By the end of 2025, its monthly revenue run rate reached $2 billion, significantly higher than the $1 billion per quarter level at the end of 2024.
Furthermore, OpenAI's preliminary 2025 financial data exhibits a classic pattern of "hyper-growth coupled with hyper-cash consumption." Annual revenue was approximately $13 billion, more than 2.5 times the $3.7 billion recorded in 2024, with a year-end monthly revenue run rate of $2 billion. This highlights the emergence of genuine commercial demand for premium ChatGPT service subscriptions, enterprise APIs, agent tools, and the developer ecosystem.
The $38.5 billion net loss cannot be simply interpreted as uncontrolled operational cash burn, as the vast majority stemmed from non-cash accounting charges triggered by a corporate structure conversion. During OpenAI's transition from a non-profit to a public benefit corporation (PBC), convertible equity previously held by investors was accounted for as a liability. Re-measurement of this liability due to valuation increases resulted in approximately $30 billion in book expenses. Total losses related to changes in the fair value of convertible equity and warrant liabilities amounted to around $41.5 billion.
After excluding these non-cash items, equity-based compensation, Microsoft compute credits, and other factors, the actual operating loss was approximately $8 billion. Therefore, OpenAI did not truly burn through $38.5 billion in cash in one year. However, even after adjusting for accounting charges, its business model remains in a phase of high-intensity loss-making expansion.
Informed sources cited by multiple media outlets indicate that OpenAI is targeting an initial public offering in the U.S. stock market, potentially as early as late 2026.
AI Demand is Robust, But OpenAI's Path to Profitability Remains Unproven
Considering the previously reported figures for 2025—approximately $34 billion in spending, around $13 billion in revenue, and a net loss of about $39 billion (with roughly $30 billion attributed to non-cash accounting charges related to the old equity structure, resulting in an adjusted loss of about $8 billion)—the core issue for OpenAI is not the absence of revenue. Instead, the challenge lies in whether its revenue growth curve can outpace the cost curves associated with model training, inference, data centers, sales expansion, and talent.
Consequently, for capital markets, an OpenAI IPO would become one of the most critical valuation tests during the AI super-cycle. On the positive side, OpenAI has demonstrated that generative AI possesses an exceptional rate of revenue ramp-up. Annual revenue of $13 billion and a monthly run rate of $2 billion provide a solid foundation for it to be viewed as a next-generation enterprise AI software platform, AI developer ecosystem, and agent entry point.
The negative aspect is that its path to profitability remains heavily dependent on several factors: a decline in computing costs, improvements in inference efficiency, increased penetration of enterprise paid services, the formation of a commercial closed-loop for AI agents, and the return on investment for data center capital expenditures.
For the AI arms race led by OpenAI, Anthropic, and Google, these figures are not a signal that AI computing demand has peaked. Instead, they indicate that leading AI application developers are raising the competitive barriers from merely "AI model training costs" to a comprehensive "AI inference war" encompassing capital, computing power, electricity, chip supply chains, and data center execution capabilities.
Initiatives like the Stargate project, involving OpenAI, Oracle Corp (NYSE: ORCL), and SoftBank Group Corp (TYO: 9984), are advancing U.S. AI data center plans to nearly 7 gigawatts of capacity, with over $400 billion in investments projected over the next three years, moving toward targets of $500 billion and 10 gigawatts.
For investors currently focused on the prospects of enterprise AI applications, OpenAI's latest performance data delivers a dual message: Enterprise AI is transitioning from pilot projects to agent-driven workflows aimed at cost reduction and efficiency gains. However, its commercialization must prove that high-frequency usage can be converted into high-quality cash flow.
OpenAI has stated that its Codex product has become one of its fastest-growing enterprise offerings, used by over 4 million developers weekly. Enterprises are accelerating its adoption for code review, automated resource provisioning, test coverage, incident response, and reasoning across large codebases within the software development lifecycle. OpenAI is also leveraging consulting and system integration partners like Accenture, Capgemini, PwC, and Infosys to accelerate Codex's integration into large enterprise workflows.
Comments