Anthropic's seven co-founders stand to gain billions as IPO nears

Anthropic was founded in 2021 by four researchers, a former journalist, a public policymaker and an engineer. Now, with trading expected to begin on Nov. 19 in what will be one of the biggest IPOs on record, each of its seven co-founders is poised to be worth many billions of dollars.

The company grew out of a defection from OpenAI. At the end of 2020, siblings Dario and Daniela Amodei were vice presidents there, but had grown frustrated with its approach to AI safety, worried that AI capabilities would advance faster than our understanding of them. They launched Anthropic early the following year, bringing along OpenAI colleagues Tom Brown, the research engineering lead on GPT-3; Jack Clark, a policy director; Jared Kaplan, a contract research consultant; and Sam McCandlish and Chris Olah, both members of the technical staff.

Jeffrey Ladish, who joined Anthropic as a contract safety engineer in September 2021 and now serves as executive director of AI safety institute Palisade Research, said the founders were not hoping to get rich but wanted to shape something important because they thought others might do it worse. He also cautioned: "It's extremely hard not to be influenced by your incentives." Anthropic did not respond to requests for comment.

Collectively the co-founders own about 10% of Anthropic, roughly 240 million class A shares apiece, or about a 1.4% stake. At the $58.90 per share price of the most recent funding round, those holdings were worth about $14 billion each. If the IPO hits its target of $70 to $78 per share, valuing the company near $1.4 trillion, the stakes would rise to $17 billion to $18 billion apiece.

CEO Dario Amodei, 43, holds the most shares: 245,259,100 including voting proxies. He earned a doctorate in computational neuroscience from Princeton in 2011, worked at Stanford, Baidu and Google Brain, and joined OpenAI in July 2016, where he co-invented reinforcement learning from human feedback (RLHF) with Brown. He has said he got into AI to tackle the "almost intractable" problems of biology and founded Anthropic because other companies were not run in a way oriented toward real impact.

Daniela Amodei, 39, is president and board chair. A fellow San Francisco native who grew up in the Mission District, she studied English literature and politics at UC Santa Cruz, graduating in 2009. She began in politics, working for Pennsylvania Democrat Matt Cartwright, then spent five years in risk management roles at Stripe.

Tom Brown, 39, chief compute officer, is perhaps the most entrepreneurial of the group. After earning a Master of Engineering from MIT in 2010, he founded startups MoPub, a mobile advertising business acquired by Twitter in 2013, and Grouper, a Y Combinator-backed dating app. Self-taught in machine learning despite a B- in linear algebra, he led GPT-3's training at OpenAI, a leap from 1.5 billion to 175 billion parameters.

Jack Clark, 37, head of public benefit and leader of the Anthropic Institute think tank, left journalism for OpenAI in 2016. The Brighton, U.K. native wrote for ZDNet UK, The Register and Bloomberg, and told a colleague at the time that he had "never believed in something strongly before" and needed to join OpenAI "and live with the consequences, which I hope are good."

Chief science officer Jared Kaplan, 42, remains an associate professor of physics and astronomy at Johns Hopkins. The Harvard-trained theoretical physicist co-pioneered scaling laws with Dario Amodei and pioneered "constitutional AI." Chief architect Sam McCandlish, 36, a Stanford PhD who researched quantum gravity, came to OpenAI as a safety fellow and led early work on Codex. Chris Olah, 34, heads interpretability research; he dropped out of the University of Toronto after one year, won a Thiel Fellowship in 2012 and reached OpenAI in 2018 via Google Brain. Both McCandlish and Olah have said they never wanted to found a company, with McCandlish calling it "our duty."

Although the founders own only about 10% of total equity, Founder LLC, which holds a single Class F share, gives them 50.1% majority voting control. Benjamin Mann, an Anthropic technical staff member and fellow OpenAI alum often described as a co-founder, is not an owner of that entity. The seven plan to collectively sell over 20 million shares in the IPO, yielding roughly $250 million in cash each, according to the prospectus. They intend to dedicate 80% of their personal Anthropic wealth to philanthropy, a pledge that is not legally binding, and the company is setting aside 5% of its shares for a new public benefit fund to finance nonprofits, research and potentially direct community grants.