Dario Amodei: How a Princeton Biophysicist Came to Run Anthropic

Dario Amodei is the co-founder and chief executive of Anthropic, the company that makes the Claude AI models. He started it in early 2021 with his sister Daniela and five other former OpenAI staff. On October 26, 2026, Anthropic filed with the SEC to list its Class A shares on Nasdaq under the ticker ANT. The filing showed $16.3 billion of revenue in the first half of 2026, against $4.6 billion in all of 2025.

He trained as a scientist. His PhD at Princeton was in biophysics, and his thesis looked at how brains process information. He then worked on AI research at Baidu, Google and OpenAI before he left to start a company built around AI safety.

Physics at Stanford and a Biophysics PhD at Princeton

Amodei began college at Caltech and finished a bachelor’s degree in physics at Stanford in 2006. He then went to Princeton for graduate school. From 2006 to 2011 he worked across the physics and molecular biology departments under neuroscientist Michael Berry and physicist William Bialek.

In Berry’s lab he studied retinal cells. He built a multi-electrode array for recording from many cells at once and later worked on pipettes that measure electrical activity inside single cells. His 2011 dissertation was about how the brain handles data. He held a Hertz Fellowship, a National Defense Science and Engineering Graduate Fellowship and a Princeton Centennial Fellowship. Berry later called him “the most talented graduate student I’ve advised while at Princeton.”

After the PhD he did a postdoctoral fellowship at the Stanford School of Medicine with Parag Mallick, researching ways to detect cancer. That work ended around December 2014, when he moved into AI full time. Students considering the same path can compare programs in our list of US universities offering master’s degrees in AI.

Dario Amodei’s Career Timeline

The OpenAI Years and the Work on GPT-2 and GPT-3

Amodei joined OpenAI in 2016 after a short stint at Google. He first led its AI safety team and rose to vice president of research by the end of 2019. In that period he led work on the large language models GPT-2 and GPT-3 and helped develop reinforcement learning from human feedback, the method that trains a model on people’s ratings of its answers. He is also known for early work on scaling laws, which describe how models improve as data and computing power grow.

He resigned at the end of 2020 after disagreements over how OpenAI was approaching safety and over the company’s direction. Several colleagues left with him. OpenAI has since gone its own way, as our coverage of OpenAI’s GPT-6.1 Terra release shows.

Founding Anthropic as a Public Benefit Corporation

Anthropic was founded in 2021 by seven former OpenAI employees: Dario Amodei, Daniela Amodei, Tom Brown, Chris Olah, Jared Kaplan, Jack Clark and Sam McCandlish. Dario became CEO. Daniela is president and chair of the board.

The company was set up as a public benefit corporation. Its legal purpose is the responsible development of advanced AI for the long-term benefit of humanity. That choice shaped the share structure in the IPO filing.

The seven together hold about 1.69 billion shares, or about 10% of the company. Each holds roughly 1.4% on a fully diluted basis.

How Anthropic’s Valuation Grew Under Amodei

Anthropic raised money in large, fast rounds through 2025 and 2026. The table shows the three most recent rounds before the IPO filing.

By our calculation, the valuation rose about 5.3 times in roughly nine months between the Series F and the Series H. The Series H price works out to $58.90 a share after a 10-for-1 stock split on September 17, 2026. At that price, Amodei’s 240.6 million shares were worth about $14.2 billion. Counting the 4.6 million shares in donor-advised funds, the total is about $14.4 billion.

Run-rate revenue, which annualizes the latest month, was $14 billion in February 2026 and $47 billion in May 2026. The SAFE notes and convertible instruments from early rounds now show up as large paper losses, which our explainer on how SAFE notes work helps decode.

Anthropic’s Finances in the S-1 Filing

We worked out one ratio from these figures. First-half revenue of $16.3 billion against about 5,300 staff gives about $3.1 million per employee for six months, or more than $6 million a year. Few software companies of any size come close.

Large customers often use Claude to answer questions from their own documents, a setup explained in our guide to retrieval-augmented generation.

Amodei’s Pay, Stake and Voting Power

Amodei’s total pay for 2025 was $17.96 million, including a salary of $654,167. Daniela Amodei’s was $16.4 million. In July 2026 both of their salaries were doubled to $1.4 million a year. Most of his wealth sits in his shares, which dwarf his pay.

Control works through a separate structure. A single Class F share, held by an LLC owned by the seven cofounders, carries 50.1% of the vote on major corporate matters and no economic rights. A cofounder stays in that LLC while holding at least 2% of the Class A shares they held at the IPO. The Long-Term Benefit Trust, whose trustees hold no equity, elects four of the seven board directors. Former Federal Reserve chair Ben Bernanke joined the Trust in July 2026.

Amazon and Google, Anthropic’s largest cloud partners, will hold non-voting Class N stock. Google’s stake is capped at 14.99%. Google has also agreed to supply Anthropic with up to 1 million of its AI chips, a link that runs through our profile of Sundar Pichai at Alphabet.

The 80% Giving Pledge and the Public Benefit Fund

Each cofounder has pledged to give 80% of their personal Anthropic equity to charitable causes. The pledges are voluntary and revocable, and they are not legally binding. Timing, amounts and recipients have not been decided. The money could go through donor-advised funds, charities or private foundations.

The company also set aside shares for giving. The Anthropic Public Benefit Fund holds 846.8 million shares, worth about $49.9 billion at the Series H price. Jack Clark is expected to be one of its first managers.

There is a governance side to the pledge. As cofounders give shares away, their economic stake shrinks, but their votes through the Class F share stay the same. Anthropic lists that gap as a risk for other shareholders.

The 2026 Dispute With the Department of Defense

The largest public fight of Amodei’s tenure came in early 2026. Amodei had said that Claude would not be used for autonomous weapons or to surveil American citizens. The Department of Defense took the view that the military, and not its contractors, decides how purchased tools are used, and Anthropic declined to drop its limits.

The department then labeled Anthropic a “supply chain risk,” and President Trump ordered federal agencies to stop using Claude. Anthropic sued, arguing the moves were illegal retaliation for its safety stance. On March 26, 2026, Judge Rita F. Lin of the US District Court for the Northern District of California issued a temporary injunction. She wrote that the designation was likely “contrary to law and arbitrary and capricious.” Microsoft, the ACLU and retired military leaders filed briefs supporting Anthropic.

Government sales have resumed. In October 2026 Anthropic signed a OneGov deal with the General Services Administration that gives federal, state and local agencies 40% off token-based Claude usage. Offices adopting Claude can start with our AI acceptable use policy template.

Amodei’s Essays on AI Benefits and Risks

Amodei writes long essays that set out his views on where AI is heading. “Machines of Loving Grace,” published in October 2024, describes how advanced AI could improve human welfare. “The Adolescence of Technology,” published in January 2026, sorts the risks of AI into five categories.

The two essays read as a pair. One makes the case for building advanced AI, and the other lists what could go wrong. That tension runs through Anthropic itself, which warns in its own IPO filing that advanced AI could pose catastrophic risks if misused.

Where Anthropic Stands as It Prepares to List

Anthropic’s S-1 left the price range and share count blank. Goldman Sachs and Morgan Stanley lead a group of 22 underwriting banks. The cofounders face a standard 180-day lockup after the listing.

The company now competes with OpenAI, Google and a growing group of other model makers, including enterprise-focused firms like Cohere, covered in our report on Canada’s investment in Cohere. It reached its first operating profit in the second quarter of 2026. The open questions for Amodei are computing costs, which rose to $5.6 billion in the second quarter, and how public shareholders will live with a company whose founders and trust keep control. Our Anthropic news roundup tracks the filing and the company’s latest announcements.

Questions People Ask About Dario Amodei

Who Is the CEO of Anthropic?

Dario Amodei has been CEO since he co-founded the company in 2021. His sister, Daniela Amodei, is president and chair of the board.

Where Did Dario Amodei Study?

He started at Caltech, earned a B.S. in physics from Stanford in 2006 and completed a PhD in biophysics at Princeton in 2011. He then did a postdoc at the Stanford School of Medicine.

How Much of Anthropic Does Dario Amodei Own?

He held 240,637,980 shares on September 30, 2026, plus about 4.6 million in donor-advised funds, roughly 1.4% of the company. With the other cofounders he also controls 50.1% of the vote through a single Class F share.

Why Did Dario Amodei Leave OpenAI?

He resigned at the end of 2020 after disagreements over OpenAI’s approach to AI safety and its direction, and co-founded Anthropic with six colleagues in 2021.

Records and Filings Used in the Dario Amodei Profile

• The Daily Princetonian: Inside Dario Amodei’s PhD at Princeton (October 2026)
• World Economic Forum: Dario Amodei profile
• Clay: Anthropic CEO career record
• Fortune: Anthropic’s cofounders, stakes and giving pledge from the S-1 (October 26, 2026)
• The Neuron: Dario Amodei’s pay and the Long-Term Benefit Trust
• Anthropic: S-1 registration statement announcement (October 26, 2026)
• INDmoney: Anthropic Series H at a $965 billion valuation
• Outlook Business: Anthropic raises $30 billion at a $380 billion valuation (February 2026)
• NPR: Judge temporarily blocks the administration’s Anthropic ban (March 26, 2026)
• ExecutiveGov: GSA and Anthropic OneGov discount
• Euronews: Alphabet’s chip deal with Anthropic (January 13, 2026)
• Dario Amodei: Machines of Loving Grace (October 2024)
• Dario Amodei: The Adolescence of Technology (January 2026)