Anthropic's IPO Filing: What the S-1 Tells Us So Far

Updated October 28, 2026

Anthropic, the company behind Claude, is moving toward the public markets, and the paperwork is now out in the open. Here is what the filing and the reporting around it tell us, and what is still unknown.

How we got here

Anthropic announced on June 1 that it had confidentially submitted a draft Form S-1 to the SEC, which gave it the option to go public once the review is finished. That came days after it disclosed a $965 billion valuation following a $65 billion funding round, a figure above OpenAI's $852 billion valuation from March. The S-1 has since been posted publicly on the SEC's EDGAR system, and news outlets are now working through its contents.

The numbers

The Wall Street Journal reported about $16.3 billion in revenue for the first half of 2026, alongside a net loss above $80 billion. Separately, the Financial Times reported that Anthropic told investors its third-quarter revenue reached $16.8 billion, up from $11.5 billion in the second quarter, and that the company is profitable on an adjusted basis. A massive reported loss next to an adjusted profit is the critical detail to watch. The likely drivers are compute costs and non-cash items, and the filing's notes should show exactly how much each contributes.

The risk factors

Reuters, which reviewed the prospectus, noted that roughly 80 of the 261 pages in the main filing are devoted to risk factors. In the filing, Anthropic acknowledges that more capable models and wider use cases could raise the risk that its systems cause harm. Analysts quoted in coverage say this creates real tension for a company that will now answer to shareholders every quarter while also saying its technology carries serious risks.

Who controls the company

Anthropic remains a Delaware public benefit corporation, meaning it has a legal duty to weigh public benefit alongside profit. The filing describes a multi-tier share structure. Ordinary Class A shares carry one vote each, while other classes held by founders and a trust-linked group carry extra influence. Earlier reporting from Reuters said the trust's trustees would be able to elect a majority of the board. Dual-class structures are common among founder-led tech companies, but pairing one with a legally binding public-benefit mission is new for public investors. Dario Amodei reportedly owns only about 2% of the company's equity, making the multi-tier voting arrangement crucial. By leveraging a trust-linked group with extra voting rights, Amodei and the founders maintain operational control of the company despite their low equity stake.

Timing and size

Reports point to a possible November listing. Wall Street expectations reported earlier by Reuters were as high as a $100 billion raise at about a $2 trillion valuation. Those are expectations, not offering terms. Share count, price range and listing date depend on market conditions.

Why it matters

• For investors: it is the first time the public can see audited details of a leading AI lab's revenue, costs and commitments.
• For customers: the filing shows how dependent the company is on large accounts and on its infrastructure partners.
• For the industry: analysts see it as a major test of investor appetite for massive AI capital expenditures, especially when contrasted with the private funding trajectories of rivals like OpenAI, setting the stage for other large tech listings expected soon.

What to watch next

1. The gap between reported and adjusted results.

2. The final offering terms: price range, share count and exchange.

3. How the governance structure holds up once public shareholders can vote.

Ultimately, this IPO will serve as a massive temperature check for the broader generative AI sector's financial viability in the public markets.

Disclosure: Anthropic makes Claude, the AI assistant used to help research and draft this post. Check key figures against the SEC filing before relying on them.