MW The real AI bubble may be in the private market

By William Gavin

xAI, OpenAI and Anthropic have been rapidly raising more money, adding to the notion of 'aspirational' valuations across the AI startup scene

OpenAI is expected to raise funds at a $750 billion valuation - 50% above its previous, $500 billion estimated valuation.

The jury is still out on whether creators of popular artificial-intelligence chatbots will be able to work out business models that justify their hefty investments. But that hasn't stopped those companies from racking up ever more money in the private market.

Valuations keep rising for some of the biggest AI unicorns, as evidenced by xAI's announcement Tuesday that it had raised $20 billion in a Series E round that exceeded its $15 billion target. CNBC in November reported that the Grok chatbot maker, launched in 2023 by Elon Musk, was expected to close a $15 billion funding round at a $230 billion premoney valuation.

That would mark a premium compared to what xAI was worth last March when it acquired Musk's social-media platform, X. That transaction valued xAI at $80 billion and X, formerly known as Twitter, at $33 billion. xAI has raised a total of $42.1 billion, according to PitchBook, which tracks private capital markets.

"It is likely that the venture market is experiencing an AI bubble. The data checks most of the boxes," Kyle Stanford, PitchBook's head of research for the U.S. venture market, said in a recent report.

"The market has shown a sharp price increase, the market narrative has become 'AI versus the rest,' and much of the investment is driven by investor envy and fear of missing out on returns," Stanford added.

As of last October, ChatGPT creator OpenAI was worth $500 billion and had raised a total $66.4 billion from investors since its inception, according to PitchBook. Anthropic ended 2025 with more than $39 billion raised since it was founded by former OpenAI employees in 2021. In September, Anthropic was valued at $183 billion, or almost three times what it was worth less than a year earlier.

Anthropic, which develops the Claude chatbot, is now in talks to raise $10 billion at a $350 billion valuation, or almost twice what it was worth as of its previous round, the Wall Street Journal reported on Wednesday. And OpenAI is planning to raise as much as $100 billion at a $750 billion valuation, according to the Journal. Earlier reports had pinned OpenAI's potential value at as much as $830 billion.

"What's really unprecedented is kind of the size of the investment going into the space right now," said Andrew Alden, vice president of quantitative research at Forge Global, which operates a marketplace where investors can trade shares of private companies.

Alden pointed to Forge's trading data, which showed that 45% of volume in 2025 was related to AI companies; that compares with 40% in 2024 and just 18% in 2023. Returns on Forge's dedicated AI basket of 19 companies grew 191% in 2025, according to Alden.

"There still needs to be a validation of the business model, and I think that is coming," he said - noting that some companies have "aspirational" valuations that are very high compared to the revenue they currently generate.

xAI recorded a net loss of $1.46 billion for its September quarter after spending $7.8 billion across the first nine months of 2025, Bloomberg News reported Thursday. And fresh off its latest fundraise, the company said this week that it will invest $20 billion to build its third data center.

xAI's revenue for the three-month period that ended on Sept. 30 amounted to $107 million. That's almost double what was recorded for its June quarter, according to Bloomberg.

Anthropic, which is reportedly a candidate for a public listing, planned to burn almost $3 billion on $4.2 billion in sales in 2025, the Journal reported in November, citing financial documents showed to investors over the summer.

See more: This crazy chart shows just how much cash OpenAI is burning as it chases AI profits

OpenAI doesn't expect to be cash-flow positive until 2030. By then, HBSC analysts have calculated that it could face a funding shortfall of $207 billion.

The company was on pace to burn $9 billion after generating $13 billion in sales in 2025 as of November, according to the Journal, which cited financial documents. Sam Altman, OpenAI's CEO, has said the company was making much more than $13 billion and that revenue was growing "steeply."

"Whether we burn $500 million a year or $5 billion or $50 billion a year, I don't care - I genuinely don't," Altman said in 2024. "It's going to be expensive. It's totally worth it."

A representative for OpenAI did not immediately return a request for comment. An inquiry sent to xAI received an automated reply. Anthropic declined to comment.

Much of what AI firms are raising is put to work developing data centers and equipping them with expensive hardware, including AI chips. Public hyperscalers, such as Oracle $(ORCL)$ and Microsoft $(MSFT)$, could collectively spend potentially $2.9 trillion on AI-related capital expenditures through 2028.

And as companies develop their AI models, they're spending more and more to train them, according to Stanford University's latest AI Index report. Anthropic CEO Dario Amodei told CNBC in early 2024 that it could cost up to $1 billion to train AI models that year. He also forecast that spending on training could rise to as much as $10 billion this year.

"I think this is the fundamental problem with generative AI and LLMs today - they are so expensive," said investor Michael Burry, of "The Big Short" fame, in a Substack debate published Friday. "It is hard to understand what the profit model is, or what any one model's competitive advantage will be - will it be able to charge more, or run cheaper?

While OpenAI, xAI and Anthropic have some of the deepest pockets among private AI firms, they're far from the only ones benefitting from investor enthusiasm.

Investment in AI accounted for more than half of the $512.6 billion deployed by venture capitalists around the world in 2025, according to data from PitchBook. A year earlier, investments in AI made up 38% of all venture-capital deal value.

"There are parts of the AI ecosystem that are probably in bubbles," Google $(GOOGL)$ $(GOOG)$ DeepMind CEO Demis Hassabis said last month. "One example would be just seed rounds for startups that basically haven't even got going yet, and they're raising at tens of billions of dollars."

Heavy investment in such startups has helped more than 50 individuals working on AI companies become billionaires in 2025, according to Forbes. Edwin Chen, whose 6-year-old business Surge AI refines AI models, is the wealthiest of the new class of AI billionaires, worth about $18 billion.

Read more: AI made tech billionaires even richer this year. Here's how much.

Former OpenAI executive Mira Murati's Thinking Machines Lab raised $2 billion at a $12 billion valuation last July in one of the largest seed rounds of all time. By November, barely a month after introducing its first product, it was in talks to raise more cash at a $50 billion valuation, Bloomberg reported.

Companies working on so-called physical AI, such as humanoid robots and driverless vehicles, have also been popular. Figure AI was most recently valued at $39 billion, while the Alphabet-backed robotaxi firm Waymo is reportedly in talks to join the $100 billion club.

Most of the top 10 venture-capital deals of the December quarter involved companies that heavily integrate AI into their businesses, according to PitchBook data. At the top of the leaderboard was Anthropic's deal with Nvidia (NVDA) and Microsoft, followed by Amazon.com (AMZN) founder Jeff Bezos's cash infusion into his physical-AI startup, Project Prometheus.

Fears of an AI bubble have also been prominent in the public markets. Stock-market bulls have argued that there's still room for companies to grow, while more pessimistic investors say that tech-company valuations are too high and could collapse.

"We see evidence of froth in circular investments between leading AI firms and suppliers, [and] concentrated valuations in a few frontier labs, while at the same time, we think the underlying fundamentals remain orders of magnitude stronger than in prior tech bubbles," Truist Securities analysts said in a recent note as they initiated coverage on several AI firms, including Palantir Technologies (PLTR).

Don't miss: Big Tech needs a staggering $1.5 trillion to fund the AI boom. This is the complex playbook it's using to get it.

-William Gavin

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

January 10, 2026 07:00 ET (12:00 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.