January 16, 2026 OpenAI could run out of money within the next 18 months. That prediction, issued by Sebastian Mallaby, a senior fellow at the Council on Foreign Relations, in a recent New York Times essay, argues that while AI as a technology is advancing at remarkable speed, OpenAI’s business model may not be built to survive the cost of winning.
The company has committed to spending well over $1 trillion before the end of the decade to train larger models and build the infrastructure needed to run them. That bet on scale has helped OpenAI stay at the forefront of generative AI, but it has also locked the company into a burn rate that few firms could sustain. Unlike rivals such as Google, Meta and Microsoft, OpenAI does not have a mature, highly profitable legacy business to subsidize those investments.
That structural gap matters. Google and Meta can funnel cash from advertising and cloud services into AI development for years if necessary. OpenAI, by contrast, relies heavily on investor capital and a consumer subscription product that has so far shown limited willingness to convert massive usage into durable revenue. ChatGPT may be ubiquitous, but most users still expect it to be free.
Mallaby is not predicting the collapse of the technology itself. He notes that “businesses usually take decades to deploy new technologies successfully,” and argues that the industry’s progress over just three years has been “striking.” His concern is narrower and more pointed in that OpenAI may not be one of the long-term winners.
That view is increasingly shared across the industry. Analysts have described 2026 as a potential “make-or-break year” for OpenAI, as infrastructure costs continue to rise and competition intensifies. Google has accelerated its Gemini rollout, Meta is building its own massive AI infrastructure and Microsoft, OpenAI’s most important partner, is simultaneously developing its own AI products.
Inside OpenAI, Sam Altman has reportedly declared “code red,” doubling down on ChatGPT in an effort to defend market share. But some investors are openly skeptical. “This is the WeWork story on steroids,” a venture capitalist told The Economist, likening OpenAI’s spending trajectory to the coworking company that collapsed after years of unchecked expansion.
For now, OpenAI remains one of the most influential companies in tech. But if Mallaby’s warning proves accurate, influence alone may not be enough to keep the company solvent.
