Dario Amodei and Anthropic's IPO: the safety lab goes public with $80,700M in half-year losses

By The Profile (Polina Pompliano) · October 24, 2026.

Anthropic could make its IPO prospectus (Form S-1) public as soon as Monday, October 26. According to IFR and Reuters, it is aiming to list in mid-November, with an investor roadshow the week of the 9th. It could raise up to $100 billion at a valuation of up to $2 trillion, which would make it the largest IPO on record. As of Saturday, there was no public S-1 in the SEC's EDGAR database.

The first figures come from leaks of the confidential draft. The Wall Street Journal reports $16.3 billion in revenue and an $80.7 billion net loss in the first half of 2026. IFR puts third-quarter revenue at $16.8 billion, up 45% from the second quarter. Potential future compute spending would be around $750 billion, according to the WSJ. These are figures from a draft, not audited, published accounts.

For a sense of scale: in May, Anthropic raised $65 billion at a $965 billion valuation (a Series H round led by Altimeter, Dragoneer, Greenoaks and Sequoia). The company itself then claimed an annualized revenue run rate of more than $47 billion. A jump to $2 trillion would require the market to pay roughly twice what private investors paid a few months ago.

The piece is a profile of Dario Amodei, co-founder and CEO. He holds a PhD in biophysics from Princeton and was vice president of research at OpenAI, where he worked on GPT-2, GPT-3 and reinforcement learning from human feedback (RLHF, the technique that made chatbots usable). He left at the end of 2020 over disagreements on safety and direction, and founded Anthropic in 2021 with his sister Daniela and five other former colleagues. According to the profile, his father died of an illness that years later became largely curable, which pushed him toward biology and later toward AI as a scientific accelerator.

Amodei is unusual among the sector's chief executives. According to the profile, he has put the probability of a catastrophic outcome from AI at between 10% and 25%, and has warned that it could eliminate half of entry-level white-collar jobs within one to five years. In September 2026 he called for the industry, his own company included, to deliberately slow the pace of model improvement. Sam Altman responded that he agreed on the need for “pacing the frontier.”

The profile also covers the criticism. In February, Anthropic refused to lift Claude's restrictions on domestic mass surveillance and fully autonomous weapons. The Department of Defense labeled it a “supply chain risk”; a judge called this “classic First Amendment retaliation” and blocked it, although an appeals court upheld the Pentagon's narrower power to exclude Claude from defense work. In addition, according to the profile, some critics argue that its safety commitments have been softened by competitive pressure. On October 14, New York City Comptroller Mark Levine asked it to pause the IPO until safety and governance issues are resolved; according to Bloomberg Law, Anthropic has opened talks with his office.

**Our take.** The thesis: with the IPO, safety stops being a brand narrative and becomes a valuation factor, and therefore a source of financial pressure. A lab that loses $80.7 billion in six months and commits to hundreds of billions in compute needs to grow very fast to justify its price. That clashes head-on with asking the industry to slow down. Both can be sincere at once, but a public shareholder is not a patient private investor.

The most interesting contrast is Amodei with himself. He champions an AI capable of curing diseases like the one that killed his father and, at the same time, estimates a double-digit catastrophic risk. That tension is, at heart, our own editorial line: a hard transition in the short term, abundance and health in the long term. But it is worth separating what has been demonstrated from what is aspirational. Growing revenue is a reported fact; the company's ability to hold its limits under quarterly pressure has yet to be proven.

What to watch when the S-1 is public: which safety and governance commitments are listed as risk factors, who outside the company can verify them, and whether the control structure shields those decisions from shareholders. In general, credible pledges of self-restraint are those a third party can check, not those the issuer declares.

The co-founders' pledge to donate 80% of their shares, and Forbes' estimate of around $15.5 billion in net worth for Amodei (with a stake of roughly 1.6%, an unaudited figure), add color but do not settle the essential question. What matters for the long term is not how much the company is worth on its debut day, but whether the money of thousands of small savers ends up pushing or curbing prudent decisions. That is where the credibility of the “safety-first” model will be decided.