Co-founder and CEO of Anthropic Dario Amodei looks on as President Donald Trump speaks to the press after a meeting with technology executives about artificial intelligence at the White House in Washington, Sept. 29, 2026.
Kent Nishimura | AFP | Getty Images
Anthropic picked a tough week to unveil its IPO prospectus.
Within days of its release, several of the company’s key partners, including Microsoft, Google and Amazon, reported quarterly results that added fuel to existing market jitters around the artificial intelligence buildout.
“Everybody’s still full steam ahead on AI, but all the things around it took an interesting turn this week,” Daniel Newman, CEO at Futurum Group, told CNBC in an interview on Friday. The hyperscalers’ results have “raised eyebrows” and created a new headwind in the market, which is “more doubt on capital structure,” he said.
No matter the current state of the market, Anthropic was going to have to address a skeptical investor base as it pursues a potentially record-setting initial public offering in the coming weeks. In its prospectus on Monday, the AI company disclosed a net loss for the first half of the year of $80 billion, alongside hundreds of billions of dollars’ worth of compute commitments and more than 90 pages of risk factors. Investors are still waiting to see Anthropic’s third-quarter financials.
Heading into this week, fears were mounting about the financial health of neoclouds like CoreWeave and Nebius, which are facing rising borrowing costs and a dearth of lenders as they pursue aggressive AI buildouts. Anthropic gets much of its capacity from Amazon, Google and SpaceX, but it also relies on neoclouds for some compute as it races to expand as quickly as possible.
Now investors are growing increasingly worried about the whole AI trade.
“It’s a great technology, seems to do things that we never would have imagined, but how does that then translate into a viable business?” David Hsu, a professor at the University of Pennsylvania Wharton School, said in an interview.
The Nasdaq Composite notched its first three-week losing streak since March after plunging 2.9% on Friday. Nvidia, whose graphics processing units have powered the AI boom, traded lower, along with the whole semiconductor sector, after Intel issued a weak forecast, sending its stock to its steepest drop since the 1970s.
Microsoft, Google and Amazon all had notes of concern this week in their reports. The companies have been on a spending blitz in recent years to meet the surge in demand for AI infrastructure, but all three issued capex figures or forecasts that came in below analysts’ estimates. Microsoft lowered its capex forecast for the year to $170 billion from $175 billion, while Alphabet maintained its guidance for up to $205 billion and Amazon reiterated its roughly $220 billion forecast.
“It’s been an ugly reporting season for tech thus far,” Jonathan Krinsky, managing director at BTIG, wrote in a Thursday note. He described the fallout as “The Techsas Chainsaw Massacre,” adding that there’s too much weakness in the sector for the market to ignore.
Cash is becoming an issue. Amazon reported negative free cash flow for the second consecutive quarter, recording an outflow of $42.4 billion for the trailing 12-month period. Google reported negative free cash flow of $9.1 billion, one reason the stock tanked almost 13% on Thursday, its worst day ever.
Microsoft maintained positive free cash flow of $18.6 billion, but it’s down significantly for a second straight quarter. The company announced plans on Thursday to raise $25 billion in bonds, as it seeks additional sources of funding to help finance its infrastructure expansion.
“Suddenly, they’re like, ‘We can’t keep funding this on free cash flow,'” Newman said. “The worry is that basically now there’s no company that’s able to do this using their own business capabilities to create cash flow. It’s becoming increasingly alarming.”
Cloud growth remains strong
Anthropic’s hyperscaler partners are also among its biggest investors. The companies all need each other to keep the flywheel turning, a topic that’s sure to be a major point of discussion in Anthropic’s upcoming roadshow.
Anthropic didn’t immediately respond to a request for comment.
A big tail wind for Anthropic remains AI demand.
Microsoft, Google and Amazon all reported cloud growth that topped analysts’ estimates, with revenue in Google’s business more than doubling from a year earlier. And Oracle, one of the biggest spenders on AI data centers, said at its investor day on Wednesday that it will limit additional fundraising in the coming years to $20 billion, and will turn cash flow positive in the second half of next fiscal year.
Analysts at Citizens wrote in a Thursday report that the announcement “put the biggest investor fear to rest — that funding its enormous AI infrastructure buildout would require an open-ended increase in debt and equity issuance.”
Attention will now turn back from hyperscaler earnings to Anthropic, as investors assess how much risk they’re willing to assume and at what price.
The IPO market has largely been frozen since SpaceX’s record offering in June. Biometric ring maker Oura, Australian AI data center operator Firmus and SK Hynix’s U.S.-based Solidigm subsidiary have all recently postponed or abandoned plans to IPO. Nscale, a neocloud that filed to go public in September, hasn’t updated its prospectus in weeks, nor has SoftBank’s SB Energy.
With Anthropic, Newman said the FOMO — fear of missing out — on a potentially “generational” company is real, and that the prospectus was “very impressive,” showing 1,300% annual sales growth, 900% net revenue retention and an implied gross margin of 85%.
“Anthropic is the IPO market at this point,” Newman said. “I think companies would have to be extremely cautious, if not just foolish, to actually want to come out in this period.”
— CNBC’s Jordan Novet, Annie Palmer and Samantha Subin contributed to this report.
WATCH: Anthropic’s heavy AI spending could hold the key to its profitability: Ark Invest
