Sam Altman’s OpenAI is facing a unit economics crisis that is becoming increasingly difficult to dismiss as mere growing pains. According to data from The Information, the company’s adjusted operating margin plummeted to minus 122% in the first quarter of 2026. To put it bluntly: OpenAI is subsidizing its dominance, burning $1.22 for every dollar it earns. Even with quarterly revenue reaching $5.7 billion, this model looks less like a market-capture strategy and more like chronic scaling inefficiency funded by a continuous stream of investor cash.
While OpenAI attempts to squeeze revenue from enterprise sales, its Codex AI agent, and initial advertising tests within ChatGPT, audience growth has begun to stall. The platform attracted 905 million weekly active users, falling short of the coveted one-billion mark. With R&D and computing costs remaining fixed at astronomical levels, any missed KPI directly threatens the company’s path to break-even. The "scale at any cost" strategy appears to be hitting both physical and financial limits, turning every new product launch into a deeper debt trap.
In stark contrast, Anthropic’s performance serves as a biting critique of its heavyweight rival. CEO Dario Amodei has managed to build a far leaner and more aggressive commercial engine. Anthropic’s annualized revenue has already hit $45 billion, leaving OpenAI and its $30 billion run rate in the position of the underdog. Furthermore, Anthropic expects to record $11 billion in revenue for the second quarter, projecting an operating profit of $600 million. While Altman is left explaining negative margins to his board, his chief competitor is proving that large language models can actually generate profit rather than just incinerating capital.
This divergence is reaching a critical point ahead of OpenAI's planned IPO in the fourth quarter. Public markets are rarely as forgiving as deep-pocketed venture capitalists; they seldom tolerate negative unit economics paired with slowing user growth. Unless OpenAI finds a way to radically slash operating costs, its public debut risks becoming an attempt to sell investors the privilege of footed the bill for Nvidia's server farm invoices.