In Silicon Valley, there’s a number that keeps interrupting people in the middle of their sentences. $852 billion. That is the current valuation of OpenAI, a company that began eleven years ago as a nonprofit research lab, operated on donations and good intentions, and is now being discussed alongside Apple and Microsoft. The figure is truly astounding. And a growing number of those who contributed to its placement are questioning its viability.
By most accounts, the $122 billion funding round that OpenAI closed at the end of March 2026 was the biggest in Silicon Valley history. SoftBank, Amazon, NVIDIA, Microsoft, a16z, Sequoia, BlackRock, Temasek, and a list of other companies that resembles the attendance sheet for a financial industry conference were among the nearly unbelievably large number of participants in the round. The business raised more than $3 billion from individual investors alone after expanding access to them through bank channels for the first time. It was an exceptional capital raise by all standards. However, concerns about what precisely $852 billion is pricing in began to surface almost immediately.
When the Financial Times revealed that some of OpenAI’s own backers were opposing the valuation, particularly as the company started shifting its product roadmap toward enterprise clients, the tension became apparent. By its own admission, OpenAI has changed its strategic priorities twice in the last six months, first in response to pressure from Google and again when Anthropic started to close the gap in a way that clearly unnerved investors. In a direct statement to the Financial Times, one early backer asked, “You have ChatGPT, a platform with a billion users growing at 50 to 100 percent annually—why are you talking about enterprise and code?” It was the kind of remark about a company you recently assisted in valuing at almost $1 trillion that is rarely made in public.
While OpenAI’s strategy debates have been taking place, it’s difficult to ignore what Anthropic has been doing in the background. The company is currently trading at implied valuations close to $800 billion on secondary markets, up from just $4 billion in early 2023. According to reports, its revenue, which was mostly driven by coding tools, increased from $9 billion to $30 billion in a seemingly brief amount of time. According to some industry observers, Anthropic’s revenue growth rate will surpass OpenAI’s in a matter of months. Investor psychology is being seriously impacted by the possibility of that happening, whether or not it does.
Sarah Friar, the CFO of OpenAI, has firmly refuted the story of internal discontent. She has claimed that the $122 billion raise was oversubscribed and finished in record time, and that assertion is difficult to refute. It’s possible that the reported skepticism is limited to a vocal minority while the majority of investors continue to express support. Measuring these things from the outside is challenging. However, the fact that the discussion is taking place at all—publicly, with named sources and documented board-level conversations—indicates that the pressure is significant enough to demand attention.
The disparity between OpenAI’s size and profitability is what makes its circumstances especially peculiar. Although the company’s monthly revenue of $2 billion would have seemed unattainable three years ago, it is still extremely unprofitable, burning through infrastructure investments, talent costs, and compute costs at a rate that makes the path to sustainable margins genuinely uncertain. A faint echo of history can be heard here. While creating something that eventually produced enormous returns, Amazon was unprofitable for years. While investors debated whether Tesla was visionary or reckless, the company spent a lot of money. Whether OpenAI falls into that category or something else entirely—a business whose cost structure might not compress the way optimists are assuming—is the question.
An additional level of complexity is introduced by the IPO discussion. OpenAI may pursue a public listing as early as late 2026, according to reports. A successful IPO would probably require the company to price itself above $1 trillion in the public markets at a post-money private valuation of $852 billion. Only a small number of companies in history have crossed this threshold, and they were all profitable when they did. That does not imply that it is not feasible. Investors have demonstrated a remarkable willingness to value AI companies based more on their potential than on their current fundamentals. However, this indicates that there is little room for error, and the upcoming months will be crucial.
Observing all of this gives me the impression that OpenAI is going through a truly challenging phase, one in which the magnitude of its success has raised expectations that are difficult to meet or turn down. One of the most popular software programs ever created is ChatGPT. The company’s global distribution, model development pipeline, and compute infrastructure are all genuinely impressive. Honestly, it’s still unclear whether any of that justifies $852 billion in the end, and it’s better to wait and see rather than rushing to a conclusion.
