Anthropic Files for IPO on Nasdaq Under Ticker ANT

Anthropic filed a Form S-1 registration statement with the U.S. Securities and Exchange Commission on October 26, 2026, taking the formal first step toward a stock market debut on Nasdaq under the ticker symbol ANT. The filing is the first time Anthropic has disclosed its full financials publicly, and the numbers are extreme in both directions: explosive revenue growth alongside one of the largest reported losses of any company heading into an IPO, most of it a non-cash accounting effect rather than cash actually leaving the business.

What Anthropic filed, and why now

The S-1 is the document a company must file with the SEC before shares can be sold to the public, and this filing did not yet set a share count or price range — those details typically follow in a later amendment closer to the actual listing. Anthropic has been widely reported to be weighing an IPO since earlier in 2026, with its target date slipping more than once, including coverage on this site back in September when the timeline was still speculative. This filing moves the story from rumor to a legally binding disclosure: the company's own numbers, its own risk factors, and its own description of how it is run, all now a matter of public record rather than secondhand reporting.

The headline number: an $80.7 billion loss

Bloomberg's reporting on the filing puts Anthropic's net loss for the first half of 2026 at approximately $80.7 billion, against revenue of $16.3 billion for the same period — itself up sharply from $1.2 billion in revenue and a $13 billion loss in the first half of 2025. Bloomberg also reports a net revenue retention rate of roughly 900%, meaning existing customers are expanding their spending with Anthropic far faster than they are leaving, a sign of strong product pull even amid the losses.

The loss figure looks alarming in isolation, but the Financial Times reports that more than $72 billion of it comes from a non-cash accounting charge: as Anthropic's valuation has risen through successive funding rounds, the value of convertible notes it issued earlier has been revalued upward on its books, and that increase counts as an expense under US accounting rules even though no cash changes hands. The FT puts the cash-relevant piece of the loss — largely compute costs — at closer to $9 billion, a much smaller and more conventional figure for a company scaling frontier AI models.

Why capital spending looks unusually light

Bloomberg Opinion columnist Tim Culpan has flagged a detail that cuts against the industry's usual AI capex story: Anthropic reportedly spent only about $482 million on capital expenditures in the first half of 2026. That is a small number next to the hundreds of billions that rivals like OpenAI and the major cloud providers are pouring into their own data centers. The explanation is structural rather than a sign of thrift — Anthropic leases the vast majority of its compute from cloud partners rather than owning it, and Culpan notes the company carries roughly $782 billion in lease obligations, effectively shifting the capital burden of AI infrastructure onto its compute suppliers' balance sheets instead of its own.

Who is paying, and who holds the board seat

The Wall Street Journal reports that Spark Capital has invested roughly $800 million in Anthropic across six separate funding rounds, and that Spark general partner Yasmin Razavi is the only traditional venture capitalist sitting on Anthropic's board — an unusually concentrated level of VC board representation for a company of Anthropic's scale, where most investors have taken board observer seats or no formal governance role at all.

On the customer side, TechCrunch reports that as of September 30, 2026, Anthropic had more than 11,900 customers paying at least $100,000 a year, more than 1,900 paying at least $1 million a year, and roughly 250 paying $10 million or more annually — a customer base skewed heavily toward large enterprise accounts rather than a long tail of small developers.

Governance: who actually controls Anthropic after the IPO

Reuters reports that Anthropic's seven co-founders will retain 50.1% of voting control after the company goes public, through a single Class F share structure that keeps a controlling majority with the founding team regardless of how many public shares are sold. Anthropic remains organized as a Public Benefit Corporation, a legal structure that explicitly permits its board to weigh its stated mission — described in the filing, per The Verge, as an ambition to "revolutionize sectors like medicine" — against pure shareholder-return maximization. Bloomberg's Matt Levine has highlighted the filing's unusual governance mechanics more broadly, including its multiple share classes, as a notable departure from a standard Nasdaq listing's typical one-share-one-vote norms.

Forbes reports an equally unusual detail buried in the filing's succession planning: each of Anthropic's seven co-founders reportedly maintains a confidential, ranked list of up to ten potential successors, and installing a new chief executive requires approval from five of the seven co-founders — a governance safeguard apparently designed to keep leadership succession inside the founding group even decades into the company's life as a public company.

The risk factors: insurance, regulation and politics

The Wall Street Journal reports the filing runs to 94 pages of risk factors, considerably longer than a typical tech IPO prospectus. Among the more striking warnings: Anthropic reportedly tells investors that a "catastrophic" AI-related incident could leave the company unable to obtain insurance coverage at all going forward, an acknowledgment of tail risks that most software companies do not need to disclose. Axios also reports the filing warns that Anthropic's public advocacy on AI safety policy — including its reported drafting of a federal AI safety bill — could make it a target for disproportionate scrutiny or selective enforcement from the Trump administration, a political risk factor that is itself a sign of how entangled frontier AI companies have become with Washington policy fights.

What's still unconfirmed

The S-1 filed is an initial registration statement, and Anthropic has not yet set a final share price range, share count, or confirmed listing date — the week-of-November-16 target reported by the Wall Street Journal is a goal, not a locked date, and Anthropic's own IPO timeline has already shifted more than once in 2026. The ultimate valuation at pricing could also land anywhere inside or outside the reported $1.5 trillion to $2 trillion range depending on investor demand in the weeks ahead. Readers should treat all figures in this piece as reported from the S-1 and from outlets that have reviewed it, not as final, audited totals that cannot change before the stock actually begins trading.

Frequently asked questions

Related articles