Anthropic's Amended IPO Filing: $17 Billion Quarter, Growth Slowing, $895 Billion in Compute Bills

Anthropic disclosed a third-quarter revenue surge alongside a return to operating losses and a vast pipeline of future compute bills in an amended S-1 filed Monday, setting up a marketing push for its initial public offering that could begin as soon as Nov. 9.

Revenue reached $16.96 billion in the quarter, up 47% from the prior three months, according to the filing. That marked a sharp deceleration from the 144% sequential growth recorded in the second quarter. The figure was roughly 12 times the year-earlier period, and more than $5 billion above the previous quarter.

The top line came with a swing back into the red. Anthropic reported an operating loss of $567 million, compared with operating income of $974 million in the second quarter. The company attributed the loss primarily to $5.0 billion of compute secured under short-term commitments, which it said costs significantly more than comparable multiyear arrangements. Net loss narrowed to $23.29 billion from $66.3 billion a quarter earlier, a figure shaped by non-operating items tied to bonds convertible into its shares.

The commitment mountain

The filing's footnotes carry the number that has drawn the most scrutiny: about $895.4 billion in spending commitments on compute and AI infrastructure, with partners including Amazon and Alphabet's Google. Anthropic expects to incur nearly half of that before Dec. 31, 2030, with about $69.8 billion due in the 12 months following Sept. 30.

Anthropic flagged that the capacity underpinning its expansion is not entirely within its control. "Certain of our compute providers and infrastructure partners depend on third-party financing or credit support to acquire chips and develop capacity, and if such financing or support becomes unavailable or more costly, contracted capacity may be delayed, repriced, reduced, or not delivered," the company said in the filing.

Compute and infrastructure charges rose to $9.9 billion in the quarter, a fourfold increase in a year. The company also booked $2 billion in non-cash charges tied to a donation-matching program under which employees pledge equity awards and Anthropic matches them with nonvoting Class N shares.

Personnel is another heavy line. Anthropic's expenses for its 5,300 workers topped $1.7 billion from July through September — an average of more than $322,000 per worker — including $1.1 billion in stock-based compensation, according to Business Insider. Costs totaled $4.8 billion over the first nine months of 2026, and the company disclosed more than $32 billion in stock-based compensation not yet reflected in its accounting.

The Register's Kettle podcast, reviewing the 614-page filing, noted that revenues outpaced compute expenses roughly two to one for the first time in the company's history, with systems editor Tobias Mann describing the document as a mix of risk mitigation that reads like a doomsday prophecy.

Valuation slips, scrutiny rises

Some current and prospective investors are weighing a $1.5 trillion valuation, down from roughly $2 trillion, after Anthropic disclosed financial details at a recent investor event, Reuters reported. The New York Post reported that the company plans to list on Nasdaq under the ticker "ANT" and will begin marketing the IPO as soon as Nov. 9, citing Bloomberg for an expected listing before Thanksgiving.

Anthropic said it had more than 11,900 customers generating at least $100,000 in trailing 12-month revenue as of Sept. 30, up from more than 1,500 at the end of 2025. It also said it is considering tighter human oversight of some research and development decisions as its models perform a growing share of that work.

The filing's risk factors run more than 90 pages and warn that advanced models could display "self-preserving behaviors," including resisting shutdown attempts or "blackmail," and that the company is a target for individuals "seeking employment under false identities, or with ulterior or malicious motives," the New York Post reported.

Overseas, Nikkei reported that Anthropic plans to raise up to $3.5 billion in Japan as part of the listing — potentially about 60% more than SpaceX raised in the country in June, which would make it the largest fundraising in Japan by a foreign company.

The deal is landing in a hostile market for AI spending. Wall Street has grown uneasy about the debt-fueled buildout of data centers and chips, widening technology bond spreads, and renewed selling in tech stocks has highlighted investor anxiety over the boom. "I wouldn't say these results are worrisome, but I do expect some of this pressure on growth and margins to be reflected in the ultimate IPO valuation," Matt Kennedy, senior strategist at Renaissance Capital, told Reuters.

Critics are pushing harder. In a Guardian op-ed, Naomi Klein and Ilana Glazer called for policymakers to block further Big Tech IPOs, warning that the offerings from Anthropic and OpenAI carry particular risk given questions about sustained profitability and what they described as "mind-bogglingly huge amounts of debt," Common Dreams reported. Economist Jeffrey Sachs separately called for AI IPOs to be halted until Congress enacts a system of licensing, independent safety evaluation, liability and financial backstops, arguing: "A company should not disclose a risk of catastrophe to humanity and then sell its shares to the public before any public authority has the power to evaluate that risk or to stop deployment."

In a Tech Policy Press op-ed, Matt Scherer argued the SEC should block the listing, noting Anthropic holds $96 billion in cash reserves against at least $582 billion in non-cancelable commitments to five companies — Amazon, AMD, Broadcom, Google and Microsoft — and warning that retail investors and retirees would be left "holding the bag when Anthropic's bills come due."