Anthropic, PBC filed a registration statement on Form S-1 with the U.S. Securities and Exchange Commission on Monday, formally kicking off what investors expect to be one of the largest technology listings of the generative-AI boom. The company said it intends to list Class A shares on the Nasdaq Global Select Market under the ticker symbol “ANT,” though the share count, price range and valuation remain undetermined and the filing is not yet effective.
What the filing discloses
The prospectus is the first public look at Anthropic’s finances since the San Francisco lab confidentially submitted paperwork in June. For the six months ended June 30, 2026, Anthropic reported about $16.3 billion in revenue, up sharply from roughly $1.2 billion in the year-earlier period, while cost of revenue tied mainly to compute climbed to about $2.4 billion. The company also recorded a multibillion-dollar net loss in the first half of 2026 as it continues to pour capital into model training and infrastructure.
Those numbers sit alongside eye-watering forward spending commitments. Anthropic said it has approximately $417 billion in future obligations to cloud and data-center providers, with additional compute deals that could add another $125 billion to $150 billion once pricing is finalized. Partnerships with Amazon Web Services, Google Cloud and Microsoft Azure accounted for about 48% of revenue in the first half of 2026, underscoring how tightly the company’s growth is bound to a handful of hyperscalers.
Governance built for safety claims
Unlike a typical Delaware C-corporation, Anthropic is structured as a public benefit corporation and will maintain a Long-Term Benefit Trust with a special Class T share that can elect a majority of directors focused on AI safety. Immediately before the offering, the company plans to issue a single Class F share to a limited liability company controlled by co-founders Dario and Daniela Amodei and five other founders; that share will carry roughly 50.1% of voting power on matters submitted to stockholders, without economic rights.
The filing also warns that artificial intelligence could pose “catastrophic or existential risks to humanity” if not deployed with safeguards, language that will be parsed by regulators and activists even as Anthropic markets Claude to enterprises and consumers. Lead underwriters listed in the document include Goldman Sachs and Morgan Stanley.
Risks investors are being told to weigh
Anthropic flags several business-specific hazards, including tensions with the U.S. Department of Defense that could cause “material revenue losses,” concentration among a small set of cloud resellers, and the capital intensity of competing with OpenAI, Google and Meta. International revenue reached about $4.5 billion in the first half of 2026, roughly 28% of the total, as the company pushes overseas offices.
Because the registration statement is preliminary, Anthropic cannot sell shares until the SEC declares it effective. The company’s news release stressed that Monday’s announcement is not an offer to sell securities. For public-market watchers, the ANT filing nonetheless sets the clock on a roadshow that will test whether investors reward Anthropic’s safety narrative—or punish the scale of its compute tab.