Anthropic has filed paperwork for an initial public offering that could become one of the largest in Wall Street history, with shares expected to begin trading on the Nasdaq Global Select Market under the ticker "ANT" as soon as next month. The San Francisco artificial intelligence company's 335-page filing pairs eye-popping revenue growth with an unusually blunt discussion of the risks its own technology could pose.
Growth Numbers That Outpace Nearly Any Startup
According to the filing reported by the Associated Press, Anthropic's revenue exceeded $16 billion in the first half of this year, roughly 14 times what it generated in the same period of 2025. That kind of expansion is rare for a company of any size, let alone one still posting enormous losses. The filing shows an $80.7 billion net loss for the period, though the company attributes $72.4 billion of that figure to a one-time accounting charge connected to its financing agreements rather than to operating costs. A year earlier, losses totaled $13 billion.
Private investors valued Anthropic at $965 billion earlier this year, and reports suggest the public offering could value the company between $1.5 trillion and $2 trillion, though no official price range has been set. If the IPO raises close to the reported $100 billion ceiling, it would surpass SpaceX's $86 billion debut in June and put Anthropic on public markets before larger rival OpenAI completes any similar move.
A Risk Section Unlike Typical Tech Filings
Most IPO filings devote space to competitive pressure, regulatory exposure and staff turnover, and Anthropic's document covers those too. But its risk section runs to nearly 100 pages and includes warnings rarely seen in corporate securities filings: the possibility that malicious actors could use its AI models to help develop chemical weapons, that authoritarian governments could deploy the technology for surveillance or repression, and that the models themselves could eventually "pose catastrophic or existential risks to humanity."
That language is consistent with public statements from CEO Dario Amodei, who in September urged the broader AI industry to slow its pace of development so safety measures could keep up. Including such warnings in a securities filing serves a specific legal function - it discloses material risks to potential shareholders - but it also puts Anthropic in the unusual position of cautioning investors about the very product it is asking them to fund.
Governance Built to Limit Pure Profit Motives
Anthropic's ownership structure reflects that tension. The company's seven co-founders, including Dario Amodei, 43, and his sister Daniela Amodei, 39, who serves as president, would retain 50.1% of shareholder voting power through a special share class after the IPO. A separate Long-Term Benefit Trust holds authority to appoint certain directors, a mechanism intended to keep financial returns from overriding public-safety considerations. Filing figures suggest each founder could hold around 1.4% of the company, worth up to $28 billion apiece at the highest reported valuation estimate.
The offering also arrives while Anthropic is in an active legal dispute with the Pentagon, which designated the company a supply chain risk after it declined to permit unrestricted military use of its AI systems. Courts have issued conflicting rulings on related cases, and both are currently under appeal. David Hsu, a professor at the University of Pennsylvania, told the AP he views the IPO largely as a practical move to widen Anthropic's access to capital rather than a shift in strategic direction.
What Remains Uncertain
No final share price or exact offering size has been confirmed, and valuation estimates reported so far are projections rather than settled figures. The outcome of the Pentagon litigation, and how it might affect Anthropic's government and defense-related business, also remains unresolved. Investors weighing the offering will be assessing not just growth metrics but a company that has explicitly told them, in writing, that its core technology carries risks it cannot fully predict or control.