Anthropic begins countdown to IPO with prospectus filing

Anthropic has fired the starting gun on one of the largest-ever initial public offerings, laying out explosive growth, heavy losses and a daunting array of risks as it kick-starts the biggest test so far of investor appetite for the AI boom.

The developer of Claude published its highly anticipated prospectus on Monday, detailing for the first time how it plans to sustain the exponential growth that propelled it from a $183bn valuation just over a year ago to $965bn in May.

The nearly 240,000-word prospectus showed ballooning revenues for the first half of 2026 of $16.3bn, up from $1.2bn over the same period the previous year. But it did not disclose the valuation the company was targeting in its IPO.

Anthropic’s net loss in the first six months of this year was $81bn, with more than $72bn related to changes in the value of convertible notes as the company’s valuation increased. The bulk of the remaining losses came from a near-$9bn outlay on computing power and infrastructure.

The company’s operating loss was $693mn in the first half, down from $2.1bn in the same period of 2025.

Anthropic also revealed plans to aggressively increase its spending in the years to come, with commitments totalling close to $800bn over the coming decade.

Goldman Sachs is listed as the “lead left” bank while Morgan Stanley will act as stabilisation agent, a lucrative appointment overseeing the crucial first sessions of trading.

Anthropic’s founders said AI could “raise productivity across every sector of the economy”. But they also warned its models “have become better able to circumvent” systems designed to make sure they act as intended, and have been found to “manipulate or subvert” monitoring.

The company pointed to risks stemming from a small number of high-value clients that account for a large slice of its revenues. While it noted that “no single customer” represented more than 10 per cent of its revenue in the nine months to September, Anthropic said its “top five” clients accounted for a quarter of revenue over the period.

The prospectus also warned that the array of risk factors facing Anthropic could interact with one another. “Many of the risks we face are interconnected and may compound or trigger cascading adverse effects across our business, which could amplify their overall impact,” it said.

Anthropic said it would issue five different classes of shares, including Class T shares that would give its Long-Term Benefit Trust the right to appoint the majority of its board and Class F stock that would hand its seven founders 50.1 per cent of the voting power.

The San Francisco-based group is pressing ahead with the IPO at a critical juncture for the AI industry, as a furious debate rages over the safety of the technology and concerns grow over whether the increasingly debt-fuelled investment boom will pay off.

The tech-heavy Nasdaq 100, which has fallen just over 7 per cent from a record high touched on October 6, was up 1 per cent in early trading on Monday.

In a sign of the more testing backdrop, existing Anthropic investors expect the company to seek a valuation of between $1.5tn and $1.75tn. In August, several of Anthropic’s backers had said they expected the group to list at a valuation of $2tn or more.

In its prospectus, the company’s founders said the advice they had received from multiple partners was: “Never let your company be worth a dollar more than it is actually worth.”

The publication of the prospectus puts Anthropic on track to list after the US midterm elections on November 3, in which the benefits and costs of the AI boom have emerged as a central issue for voters, but before the Thanksgiving holiday on November 26.

Disclosing the prospectus is likely to ease the nerves of investors who feared that Anthropic’s window to go public was rapidly closing. It is rare for companies to list between the Thanksgiving holiday and the end of the year.

Anthropic has emerged from the shadow of arch-rival rival OpenAI this year, surpassing the creator of ChatGPT in revenues and traction with business customers.

Anthropic was founded in 2021 by a group of seven former OpenAI researchers and engineers after they became disillusioned with the approach to AI safety at Sam Altman’s AI lab.

Chief executive Dario Amodei and other AI leaders including Altman and Elon Musk have called for a slowdown in AI research to ensure the technology does not run catastrophically out of control.

Investment banks including Morgan Stanley and Goldman Sachs will set Anthropic’s valuation following conversations with investors over the coming weeks.