This year could be a pivotal one for OpenAI.
The San Francisco artificial-intelligence giant has all kinds of ambitious plans for 2026 and coming years, including developing a consumer AI device, producing its own customized chips and embarking on a massive build-out of data centers, all while keeping pace with rival AI developers.
But OpenAI’s need for more cash comes amid growing concerns about how and when the massive investment that’s been made in the company and the sector as a whole will pay off, and whether a financial bubble has formed around both.
At the end of the year, OpenAI is likely to still be the most important company in the generative-AI space, said Rob Siegel, a lecturer in management at Stanford Graduate School of Business. But there’s a lot of uncertainty about how things will play out in the near future, and there’s a risk the company could go bankrupt if investor confidence in it were to suddenly collapse, he said.
OpenAI is “a hugely important company,” Siegel said.
“It’s a once-in-a-generation company,” he said. “But that doesn’t mean it’s going to be all OK economically, given the economic model.”
“I’m not saying [a collapse] will happen,” he said. “But it’s not a crazy story.”
OpenAI representatives did not respond to a request for comment.
The company’s future is of utmost importance to San Francisco. Almost by itself, OpenAI has helped to revive The City’s tech sector from its post-pandemic slump, raising more than $60 billion and creating in ChatGPT a product that has seen one of the fastest adoption rates ever. It has also helped to boost San Francisco’s struggling commercial district, expanding its office space and workforce in The City.
And OpenAI has been crucial to San Francisco becoming ground zero for the nascent AI industry. It’s not just the premier private company in the sector and one of the most highly valued private companies in the world — it has helped to spawn a slew of other AI companies in the sector that are based in San Francisco or nearby. Chief rival Anthropic was founded by OpenAI alums, for example, as were Thinking Machines Lab and Safe Superintelligence.
OpenAI’s headquarters in San Francisco, Nov. 3, 2025.
Aaron Wojack © 2025 The New York Times Company
But OpenAI also provided plenty of fodder for its skeptics and critics. Thanks to its stepped-up investments, its losses are swelling, reportedly growing significantly larger than it previously expected. Around the beginning of September, the company projected it would burn through $115 billion in cash between then and 2029, according to The Information. That was $80 billion more than it had forecast earlier in the year.
Still, despite OpenAI’s investments, its competitors appear to be catching up in both technology and the market. GPT-5 disappointed customers, because it wasn’t as big an advance as was expected and had some of the same flaws as its predecessors.
Likely as a result of the increased competition, OpenAI’s share of the enterprise market has fallen from about half in 2023 to 27% this year, according to Menlo Ventures. And more recently, traffic to ChatGPT has dropped, falling about 5% from October to November, according to web-analytics firm Similarweb.
Yet the company is reportedly out raising money, seeking as much as $100 billion at a valuation of about $750 billion, according to The Information. That’s after seeing its valuation hit $500 billion as part of a stock sale in October. And it reportedly could be seeking a valuation of as much as $1 trillion in its prospective IPO.
AI threatens to disrupt plenty of industries and change much about how business gets done, said Siegel. So it makes sense for OpenAI to raise as much money as it can to play as big a role as possible in that disruption, he said.
OpenAI and CEO Sam Altman, left — seen speaking with Federal Reserve Board Chairman Jerome Powell in July — have committed to spending some $1.4 trillion in coming years.
Mark Schiefelbein/Associated Press
But it’s seeking additional cash at a time when investors are growing increasingly skeptical of the AI business and its data-center build-out, he said. Oracle, which is teaming up with OpenAI on Stargate, has seen its share price fall by more than one-third since November. CoreWeave, another data-center operator that’s partnered with OpenAI, has seen its share price fall even further.
The key questions hanging over OpenAI — and the rest of the sector — are how long will it take for the company to start generating positive cash flow and what is it going to take for the company to get there, Siegel said.
“OpenAI has got some real technical amazingness to it,” he said. “That works to their advantage. The biggest issue is just how much capital is it going to be until they don’t need other people’s money.”
If investors lose faith or lose hope, things could collapse quickly, he said.
Gary Marcus, a longtime skeptic of OpenAI and of generative AI technologies, echoed that concern in a December blog post. The company has likely spent nearly all of the massive amount of money it had previously raised and probably has less than a year’s worth of cash left, he said.
More than 10 years after UCSF docs started consortium of research institutions, Stanford University joins fold
The fortunes of The City’s business core were a study in contrasts — growing vibrancy and investment beside evidence of continuing economic malaise
Champagne toasts, balloon drops and skating sessions are among the ways you can bid adieu to 2025
That means it’s highly dependent on investors to keep going. And few individuals or institutions have the amount of capital OpenAI needs, he said.
Marcus, a professor emeritus at New York University and the author of “Taming Silicon Valley: How We Can Ensure That AI Works for Us,” noted in his post that he’d called out OpenAI’s shortcomings nearly two years before.
“Things took longer to unravel than I thought they would, but OpenAI’s unraveling has begun,” he wrote.
OpenAI isn’t destined to fail, Siegel said, but investor skepticism could force the company to rethink its business and commitments.
If it has a tougher time raising money, it might have to scale back on some of its ambitious plans, he said, which could affect both its business trajectory and the potential return for its current investors.
With less ready cash, “they’ve got to grow more slowly,” Siegel said. “They’ve got to invest in fewer things.”
That might not be a bad thing, said Ram Bala, an associate professor of AI and analytics at Santa Clara University’s Leavey School of Business.
Startups often are more innovative when they are operating under financial constraints, Bala said. OpenAI was arguably more innovative when it had far less money at its disposal, he said.
By contrast, when money is easy to come by — as it has been in recent years for OpenAI — it’s far easier to waste, he said. OpenAI seems to be doing that by spreading itself across so many different areas, Bala said, competing in both the consumer and enterprise markets, working on AI agents, and trying to build so-called artificial general intelligence at the same time.
“What’s happening with OpenAI is they are [in] too many different product categories and markets at the moment,” Bala said.
“You run into the issue of jack of all trades and master of none,” he said.
The bigger danger for the company and the industry is that there’s a near-term investment bubble that’s forming, Bala said.
Open CEO Sam Altman meets Larry McManus of Texas Economic Development and Tourism at the data center in Abilene, Texas on Tuesday Sept. 23, 2025.
Matt O’Brien/Associated Press
Bala said that in the long term, he’s confident there will be plenty of demand for AI-powered services. But in the short term, there’s a danger that investment in the sector will get far ahead of demand, he said. That happened at the turn of the millennium, when telecommunications companies overinvested in fiber-optic lines, he said.
As the company that has drawn far more capital and is making far more investments than any other in the sector, OpenAI is at particular risk if the bubble pops, Bala said.
Unfortunately, because so many companies are now tied to OpenAI, the fallout could be wide, he said.
“I’m more bullish about AI in general in the next one year,” he said. “I’m less bullish about OpenAI.”
To be sure, Bala said he sees plenty of opportunity for the company too. Were it to focus on the consumer market, where it has a big lead, and build on the partnerships it’s already struck with Intuit and Instacart, it could cement its place as the go-to AI service for everyday users, he said. If consumers are using ChatGPT to handle their financials, shop for groceries or perform a bunch of other tasks, they’re going to be far less likely to switch to a competing service, he said.
OpenAI’s investment in data-center capacity could also pay off with consumers by offering it the ability to deliver answers nearly instantly, Bala said. In many cases, enterprises don’t need to have that kind of immediate response, he said. But a customer who’s out shopping for something and looking for an answer isn’t going to wait for an hour for it, he said.
The investment in data centers to speed the service for consumers is “a bet, more than AGI, that would have genuine financial value,” Bala said.
For his part, Billy Riggs, a professor of management and engineering at the University of San Francisco, didn’t express any of the same worries about OpenAI as Bala, Siegel or Marcus. Instead, he said he’s bullish on the company’s future, both in the short and long term.
Unlike Bala, Riggs said he thinks it’s not OpenAI but competitors such as Meta and Google that are spreading their development among too many products. OpenAI, by contrast, is focused on developing its core AI technology, he said.
In the enterprise sector, OpenAI has established itself and has a stable core of customers, he said. And it’s created a technology that is revolutionizing the way businesses are built, allowing startups to develop money-making products quicker and with smaller teams, he said.
Riggs said he’s not worried about the money OpenAI is losing, seeing it as a necessary step to building its product and business. And he said he has no doubt that the company will find investors willing to continue to fund it, particularly in an IPO.
“I think institutional investors, particularly, really are eager to have some level of exposure to this type of potentially high-risk, high-reward product,” he said.
If you have a tip about tech, startups or the venture industry, contact Troy Wolverton at twolverton@sfexaminer.com or via text or Signal at 415.515.5594.