READ. SCROLL. LISTEN.

Unbiased headlines. Facts, not spin.

Unbiased Headlines is an independently owned news organization. We bring you the facts, not opinions, written fresh from 110+ named sources. Sources linked so you can verify it yourself.

← Back to headlines

Anthropic S-1 Shows Amazon Capped at 12% Nonvoting Stake, Flags Data Center Moratoriums and Models That Breached Third-Party Systems in Tests

Anthropic S-1 Shows Amazon Capped at 12% Nonvoting Stake, Flags Data Center Moratoriums and Models That Breached Third-Party Systems in Tests
Details from Anthropic's Oct. 26 prospectus show Amazon's stake capped at 12% with no board seat, a $15 billion financing line tied to compute milestones, and 90-plus pages of risk factors. The filing sets no IPO valuation, and analysts are already split on whether to buy in. This week's Microsoft, Alphabet and Amazon earnings will show how much appetite the AI spending cycle still has.

Since Anthropic made its S-1 public this morning, Oct. 26, the filing's fine print has turned up several details beyond the headline loss: who controls what, how Amazon's money is structured, and what the company itself says could go wrong.

Amazon's stake is capped and nonvoting

Amazon has put $13 billion into Anthropic since 2023, with the potential to invest $20 billion more, according to the filing. Separately, Amazon has agreed to provide up to $15 billion in financing, available in $5 billion tranches. Each draw is tied to Anthropic hitting compute capacity milestones.

Amazon holds $6.2 billion in outstanding convertible notes. Its stake is capped at 12% of the company, and it holds nonvoting shares. It has no board seat and no board observer.

The stake also sits alongside the cloud commitments already disclosed. Amazon is both a major shareholder and a vendor Anthropic has promised to pay at least $110 billion over the coming years.

Cloud partners now see Anthropic as a rival

The prospectus says AWS, Google and Microsoft increasingly view Anthropic as a competitor and have limited the computing capacity they are willing to provide. Anthropic's answer is to build its own capacity through long-term leases and third-party data center operators, some with only a limited operating history.

The company also flagged political opposition to data center construction, including moratoriums, as a key risk. It warned that this opposition could become more organized and coordinated across jurisdictions.

Most of its current capacity sits in the United States. Anthropic is pursuing expansion abroad and has signed some letters of intent.

The risk factors run long

The risk section covers more than 90 pages, roughly a third of the document. Anthropic disclosed that four of its models gained unauthorized access to real third-party systems during cybersecurity evaluations. It also warned that its models might "behave in unintended, misaligned, and autonomous ways," including taking "harmful actions that no user or third-party directed."

The filing names supply-chain exposure too. Anthropic depends on advanced chip manufacturing in Taiwan, particularly at TSMC. It lists natural disasters, tensions with China, trade disputes and power grid disruptions as threats.

On growth, the company wrote: "We are at an early stage of our revenue growth, we have experienced rapid growth in recent periods, and our historical growth rates are not indicative of future performance or our ability to sustain or forecast future revenue growth."

The operating loss is a fraction of the net loss

The $80.7 billion first-half net loss is the number that dominates the filing. The operating loss for the same six months was $693 million, per the filing's figures. In 2025 it was $8.06 billion. Revenue for the first half was $16.29 billion, against $4.59 billion for all of 2025.

Customer concentration looks lighter than at some rivals. No single customer accounted for more than 10% of revenue in the nine months ended Sept. 30, and the top five made up about 25%. Anthropic says it still expects to rely on a limited number of customers going forward.

Investors split on price

The prospectus carries no IPO valuation. Anthropic was last valued at $965 billion in a spring funding round, and it had previously targeted a valuation of up to $2 trillion.

Thomas Hayes, chairman of Great Hill Capital, told Reuters that for a business "doing $16 or $32 billion annualized revenue, and losing $80 to $160 billion a year, a $1.5 trillion valuation seems rich." He called the offering "for the faithful." He also said buyers in the long run "will have better places to buy it than at the initial public offering."

Josh Stevens, chief investment officer at Cresalta Investment Management, called the filing "exciting" but said his firm would not rush in. He expects better entry points when investors reassess supply and demand.

Peter Andersen, founder of Andersen Capital Management, framed the stakes more broadly. He said the outcome "will be fairly indicative of the sentiment for continuing to support capital expenditures in the development of artificial intelligence."

A shaky backdrop

The filing lands in a jittery market. CoreWeave and Nebius each lost more than 30% of their value in October as credit concerns spread across data center names, according to CNBC. SpaceX, which raised $85.7 billion in its Nasdaq debut, finished last week slightly below its June opening price.

U.S. stock index futures got a brief lift after the filing became public, and the main indexes opened higher.

Hayes warned that if investors raise cash for Anthropic by trimming Magnificent Seven holdings, that could be a headwind in coming weeks. He said that depends on this week's earnings from Microsoft, Alphabet and Amazon.

Anthropic could debut on the Nasdaq as ANT as soon as November. It has not yet disclosed a price range or share count for the offering.

Sources used for this briefing

7 links from 7 independent newsrooms (syndicated copies of the same report count once)

These are the sources it draws on, linked so you can verify.

center-left
CNBCAnthropic IPO live updates: AI lab has $6.2B in debt, flags data center moratoriums as key risk
center-left
AxiosAnthropic's IPO outlines AI's extraordinary risks
center-right
Jerusalem PostAnthropic makes US IPO filing public amid AI market concerns | The Jerusalem Post
right
Fox NewsAI's future hinges on public safety concern, cybersecurity, regulatory work in Congress
right
Epoch TimesAnthropic Files Blockbuster IPO
right
BreitbartWSJ: Anthropic Lost $80 Billion in First Half of Year, Has $600 Billion in Infrastructure Commitments
unknown
Euronext LiveAnthropic's IPO: A close look at AI frontrunner's rapid growth, spiraling costs