Anthropic is preparing to go public amid growing investor skepticism toward the AI industry. However, investors believe the company can no longer keep raising money indefinitely in the private market: its commitments to expanding computing infrastructure have reached a scale that requires access to public capital markets for financing, Axios reports.

$750 billion for computing capacity

According to a draft IPO prospectus reviewed by The Wall Street Journal, Anthropic’s commitments to fund computing capacity in the coming years could reach $750 billion.

The company has already taken on more than $600 billion in future infrastructure spending. It is now negotiating additional agreements that could increase that amount by another $150 billion.

The scale of the spending is also reflected in its financial results. In the first half of 2026, Anthropic reported revenue of $16.3 billion, up from $1.1 billion in the same period a year earlier. At the same time, the company posted a net loss of $80.7 billion. As The Wall Street Journal notes, a significant portion of that amount is tied to accounting expenses incurred in connection with raising financing.

To continue growing, Anthropic needs sums that are becoming increasingly difficult to raise solely from venture investors and private lenders. Going public would allow the company to raise funds from a much broader pool of investors.

Private financing is becoming more expensive

A telling example is the terms on which Anthropic is raising debt. According to the Financial Times, the company is offering investment group Blackstone a yield of more than 10% on $9 billion of junior debt. This is part of a record $60 billion chip-financing package.

For comparison, the yield on the previous deal was 8.5%. Rising rates indicate that private investors are demanding higher compensation for the risks associated with building AI infrastructure.

The problems are not limited to Anthropic. Borrowing costs for CoreWeave, which specializes in leasing computing capacity, have risen sharply, and the yields on some of its bonds have approached 15%.

Other troubling signals are already emerging in the market as well. Australian company Firmus, backed by Nvidia and focused on building data centers, scrapped its planned IPO amid market instability. Shares of South Korean cloud provider Elice fell 32% on their first trading day despite rapid revenue growth.

Against this backdrop, Anthropic will have to convince investors that its enormous infrastructure commitments are justified by future returns.

Why the company can’t remain private forever

OpenAI’s recent funding round of more than $120 billion shows how difficult it has become for even the industry’s biggest players to raise capital. It took months to arrange the deal, which became the largest funding round in history. In current market conditions, repeating such a result will not be easy.

At the same time, demand for Anthropic shares remains extremely high. Earlier, one potential investor even offered to trade a house worth nearly $5 million for a stake in the company.

However, strong interest in the stock does not guarantee a successful IPO. After a period of aggressive investment, the market has become more cautious about business models that require enormous spending on computing capacity and data centers. Investors now want to understand not only AI’s growth prospects, but also how companies plan to pay for that growth and when it will begin generating sustainable returns.

Going public will give Anthropic access to a deeper pool of capital than what is available in the venture market. But at the same time, public status will require greater transparency and place its financial results under close shareholder scrutiny.

IPO delayed amid concerns about AI safety

According to Axios, Anthropic had originally planned to go public in September. However, those plans changed after one of the company’s researchers departed and a series of incidents once again drew public attention to AI safety issues.

Now, as Bloomberg has reported, the company is targeting an IPO before Thanksgiving in the United States, which in 2026 falls on November 26.

The timing of the offering is far from ideal. On Wall Street, concerns are growing about the scale of investment in AI and the resilience of companies building infrastructure for the sector. As a result, Anthropic’s IPO will test not only the company itself, but also the market’s willingness to keep funding the AI boom.

One Anthropic investor interviewed by Axios believes the company has already reached the necessary scale and maturity to go public. He also described Anthropic’s latest financial results as impressive, dismissing concerns about slowing revenue growth.

Anthropic may therefore be forced to go public under less-than-favorable conditions. But for a company whose future infrastructure obligations are measured in the hundreds of billions of dollars, waiting for a calmer market also carries risks. A successful listing could restore investor confidence in the prospects of the AI industry—while a failed one would, by contrast, deepen doubts about how sustainable the economics of the current boom really are.