OpenAI is the 2025 Yahoo Finance Company of the Year

Everywhere investors looked in 2025, there was OpenAI (OPAI.PVT).

The company struck megadeals with Microsoft (MSFT), Oracle (ORCL), Advanced Micro Devices (AMD), and Nvidia (NVDA). It amassed $1.4 trillion in committed spending over the next eight years. It pushed ambitions to build its own data centers and hardware products. And it converted to a for-profit company — with a potential trillion-dollar IPO on the horizon.

OpenAI’s private company shares have surged 153% this year as it landed a $500 billion valuation, according to Yahoo Finance’s private company data.

“It’s been a year where we have really hit kind of this day one of the next phase for OpenAI,” CFO Sarah Friar told Yahoo Finance Executive Editor Brian Sozzi in an exclusive interview (video above). “And we’ve seen … a lot of the ecosystem really start to move toward us to help us create this future.”

The world’s most valuable startup is Yahoo Finance’s 11th annual Company of the Year award winner, three years after the release of ChatGPT 3.5 ushered in this generation’s gold rush. OpenAI secured the award as Yahoo Finance’s editorial team believed the company had the greatest impact on the investing narrative this year.

The Sam Altman-led firm has been at the center of the artificial intelligence boom: When it announced its multibillion-dollar deal with AMD on Oct. 6, the chipmaker’s stock shot up 24% in one day. AMD CEO Lisa Su told Sozzi at the time that OpenAI would help lay the “foundation” for the AI world.

Oracle’s stock rose 38% in the two months after it announced a $300 billion data center deal with OpenAI on July 22. CoreWeave’s (CRWV) stock more than tripled between March and October, during which it announced three deals to provide computing power to OpenAI.

“There is a lot of impact OpenAI [has] on the broader stock market,” said Rishi Jaluria, managing director at RBC Capital Markets. “I don’t want to say it’s entirely OpenAI, but there’s a lot in the stock market, even with this most recent pullback, riding on the AI trade.”

The potential growth for OpenAI is vast.

HSBC estimated that the consumer market for large language models (LLMs) will reach $129 billion in annual revenue by 2030, while the enterprise LLM market is projected to reach $386 billion.

OpenAI currently boasts 800 million weekly active users, 1 million business customers, and $13 billion in 2025 revenue. Reuters reported that the startup may go public with a $1 trillion valuation as soon as the end of 2026. Friar left the door open to that happening in her interview with Yahoo Finance. The company’s internal projection has revenue reaching $200 billion in 2030, according to the Information.

To get there, OpenAI’s spending plans are even more eye-popping. The startup has committed $300 billion and $250 billion to Oracle and Microsoft, respectively, plus $38 billion to Amazon (AMZN) and $22.4 billion to CoreWeave.

It also plans to bring major business to the chipmakers, agreeing to deploy up to 6 gigawatts of AMD’s GPUs (at an estimated cost of $150 billion for the chips alone), plus 10 gigawatts of Nvidia’s GPUs (estimated to cost $300 billion for the chips and $500 billion overall for the build-out). It has a partnership with Broadcom (AVGO) for 10 gigawatts of custom chips.

Friar defended the startup’s spending plan by citing its unprecedented growth. “The compute that we’ve used in 2025, we didn’t just find in ’25, we actually had to make a decision probably in 2023,” she said. “And what we have seen is that as we invest in compute, our revenue grows almost at the same pace.

“So if you look in 2023, the total compute we exited that year was only a mere 200 megawatts. We’ll exit this year at two gigawatts, so 10x, but our revenue, similarly, or our ARR went from about 2 billion exit in 23 to almost 20 billion now exiting 2025, actually a little bit more than 20 billion.”

And OpenAI isn’t slowing down as it races toward a world with artificial general intelligence (AGI).

“We just see no reason to stop right now in terms of the requirements and the need for compute,” Friar said.

But the gargantuan numbers have spooked investors. The startup has raised nearly $50 billion so far, but HSBC estimates OpenAI will have a $207 billion funding gap by 2030 — even if its revenue multiplies as projected.

In early November, investor Brad Gerstner asked CEO Sam Altman during a podcast interview, “How can a company with $13 billion in revenues make $1.4 trillion of spend commitments?”

Altman responded bluntly: “If you want to sell your shares, I’ll find you a buyer.” Shortly thereafter, Friar made a comment at a Wall Street Journal event suggesting that the federal government should guarantee the debt for the AI build-out, which sparked intense backlash and a subsequent backtracking from OpenAI.

“OpenAI has to scale back its ambitions, focus on the things it’s actually good at,” said Gil Luria of DA Davidson, who noted OpenAI’s LLM and chatbot are its competitive advantage.

“If you just refocused on those two things, OpenAI would be OK,” Luria added. “They started getting into trouble when they started trying to build their own data centers and build their own chips and build their own hardware, and this is why they made all these massive, massive commitments.”

That reality has sunk the stocks of its partners, as fears about an AI bubble grew in November. Oracle’s shares dropped 23% in the month, while AMD shed 15%. CoreWeave plunged 45%. Microsoft, its biggest shareholder, slid 5%.

“Oracle’s stock price is now below where it was [when it] announced the OpenAI deal,” Luria said. “This tells you that not only does nobody believe they’ll ever get paid by OpenAI, but they hurt their credibility by making it seem like they will get paid by OpenAI.” Oracle’s shares slid another 11% after it reported earnings in December.

Yet most analysts remain bullish on the future of AI, pointing to the demand top hyperscalers have reported. Nvidia, the chip supplier to the tech giants, revealed in the third quarter that it has a $500 billion order backlog for AI chips stretching to the end of 2026. A slowdown, if it comes, is likely not in the cards for next year, several experts told Yahoo Finance.

Investors have to parse through the nuances of the bubble talk, Jaluria explained.

There is likely a valuation bubble for AI stocks, a sentiment commonly shared among analysts, Jaluria said. The current spending on capital expenditures has been driven by hyperscalers and AI startups, rather than enterprise demand. Concerns surrounding data privacy and hallucinations persist, per Jaluria.

However, the long-term investment thesis for generative AI remains intact. If the models continue to become more powerful, businesses will unlock their budgets for AI spending.

“So I’m hesitant to say we’re in an AI bubble from a spending and usage and adoption perspective,” he concluded.

The first-mover advantage OpenAI once had is evaporating in 2025.

Google’s Gemini 3, released in November, has claimed several industry benchmarks. Salesforce (CRM) CEO Marc Benioff, whose company inked a partnership with OpenAI in October, posted on X that he’s smitten with the chatbot. Former Intel (INTC) CEO Pat Gelsinger also showered the model with praise.

Anthropic (ANTH.PVT), which won backing from Google and Amazon, scored $15 billion in investments from Nvidia and Microsoft and is reportedly targeting an IPO itself.

“The gap between the top models seems to be narrowing … but also the gap between the top models and everyone else is widening,” Jaluria said of the battle between OpenAI, Google, and Anthropic.

Altman declared “code red” in an internal memo earlier this month, as he redirected resources to ChatGPT, according to the Information. On Dec. 11, the company released ChatGPT 5.2 and garnered a licensing agreement and $1 billion investment from Disney (DIS).

“Code red is really just saying to the company, the main thing needs to be the main thing,” said Friar, framing the current focus as “making sure that our [ChatGPT] models are on a faster kind of speed of velocity of iteration.”

She conceded that “getting into like who’s got the best model is probably something that keeps changing.” The difference maker now can be APIs and how they companies can fully utilize chatbots’ capabilities. The team is aiming for the next iterations to have more personalization, app integrations, and “super agent” abilities, she explained. The key to OpenAI’s potential trillion-dollar IPO will be whether it can stay ahead of its rivals. Grabbing the lion’s share of growing AI spending would go a long way toward making its valuation tangible to investors.

However, the real winners of the genie that ChatGPT unleashed may be the incumbents.

Demand for cloud services has soared since 2022, boosting the coffers of Microsoft, Google, and Amazon. Though any stumbles from OpenAI will be a setback for Microsoft, it is best positioned among the startup’s major partners.

Luria hypothesized that OpenAI will prioritize its deals with Microsoft should it face cutbacks due to the software giant’s share ownership and rights over its tech.

“Everybody else may or may not get paid but will have to renegotiate and then suffer the consequences of disappointment,” he said.

The big three hyperscalers also have a wide base of customers. On the other hand, RBC estimated that over half of Oracle’s AI order backlog is from OpenAI, and the startup accounts for the majority of Oracle’s projected cloud revenue from 2027 to 2030.

Read more: What’s ahead for stocks and gold in 2026? What experts are watching.

Even more vulnerable is the newly public CoreWeave.

“CoreWeave has borrowed a tremendous amount of money to build specifically for OpenAI,” Luria said, noting the OpenAI deal represents most of its growth. The company reported $311 million of net interest expense in Q3.

As for semiconductor companies, Nvidia, with its market-leading position, has a plethora of contracts outside of OpenAI. Broadcom has other major customers like Google too.

But OpenAI is one of the first big players to bet on AMD’s next-generation AI chips, making any pullback difficult to replace.

On the upside, the credibility AMD gained by partnering with OpenAI can be used to win other customers.

“When Lisa Su took over in 2014 … they had less than half a percent market share in servers. Today, they’re 41%,” Benchmark analyst Cody Acree said of AMD’s race with Intel.

“The repeatability of that methodology, where they go in and compete on designs and compete socket for socket against Nvidia and chip away at its market share slowly and steadily over a many-year period, is a playbook that they’ve learned very well in the server space,” Acree said.

Regardless of OpenAI’s successes or failures, the AI ecosystem has grown far beyond its most famous chatbot.

“We had like a dot-com bubble and a telecom bubble … even with all of that, the way that we use the internet today is leaps and bounds above what even your most fervent bull could have dreamed of 30 years ago,” said Jaluria, who compared the current AI landscape to the dial-up era of the internet.

“I think we’re going to be incredibly surprised by how we use AI 10, 15, 20 years from now, relative to today.”

Click here for the latest technology news that will impact the stock market

Read the latest financial and business news from Yahoo Finance