Who are Anthropic's seven cofounders, the people behind Claude?
As Anthropic's record listing puts the company in the spotlight, here is who founded it and what each of the seven does today.
The short version
- Dario Amodei is CEO and sets research direction; he holds a Princeton doctorate in biophysics.
- His sister Daniela Amodei is president and board chair, overseeing the executive team.
- Jared Kaplan, the chief science officer, helped develop research on scaling laws.
- Chris Olah leads interpretability work and helped draft parts of Claude's constitution.
- Tom Brown, who led GPT-3 engineering at OpenAI, now secures computing capacity as chief compute officer.
Why it matters
If you rely on Claude for work, it helps to know who steers the company and what each leader prioritizes. The roster shows a firm built around research, safety and computing capacity, not only product launches. That is useful context while the company's listing keeps it in the news.
A founding team with OpenAI roots
As Open Magazine reported in a quick-take explainer, Anthropic was started by seven people, and the piece walks through what each one does now. It is background rather than breaking news, but it lands in a week when the company's record share sale has put its leadership under a brighter light.
A pattern shows up quickly: several of the seven worked at OpenAI before Anthropic. Dario Amodei was a vice-president of research there. His sister Daniela was a vice-president of safety and policy. Jack Clark and Tom Brown also came from OpenAI, and Brown led the engineering of GPT-3, an earlier OpenAI language model.
The two siblings at the top
Dario Amodei, the CEO, is a scientist by training. He earned a doctorate in biophysics at Princeton and did postdoctoral research at Stanford's medical school. At Anthropic he decides the research direction and overall strategy for Claude.
Daniela Amodei is president and chair of the board. Her background is unusual for the field: she studied English literature, then worked in risk management at the payments company Stripe before moving to OpenAI. Today she coordinates the executives who run research, engineering, products and commercial operations.
Put together, one sibling steers the science and the other keeps the organization running. That is a common split in young companies, but here it sits inside a firm building some of the most closely watched AI systems.
Society, science and scaling
Jack Clark is the one cofounder who came from journalism. He covered technology for Bloomberg, then joined OpenAI in 2016. He now heads public benefit at Anthropic and leads an institute that studies how AI affects society. His brief is to ask what more capable systems mean for people outside the companies that build them.
Jared Kaplan is a physicist by training and has worked in particle physics and cosmology. He helped produce research on scaling laws, which are observed patterns in how an AI model's ability changes as you give it more computing power, more data or a bigger design. He holds the title of chief science officer and directs the technical research, including safety work.
Scaling laws matter because they gave labs a way to predict what spending more would buy. That prediction is part of why the industry has poured so much money into computing capacity.
Computing, training and looking inside
Tom Brown is chief compute officer. His job is to line up the computing resources needed to train and run Claude, and to make sure the company uses them efficiently as demand grows.
Sam McCandlish was previously the chief technology officer. He became chief architect after Rahul Patil joined as CTO. He concentrates on training models at large scale and on the infrastructure researchers use for reinforcement learning, a method that guides a model's behavior by rewarding the outcomes you want.
Chris Olah heads interpretability research, which tries to understand what happens inside a model that leads to a particular answer. He also helped write the parts of Claude's constitution that deal with its nature, identity and psychology. The challenge he works on is a central one in the field: even builders often cannot explain exactly how a system reached its output.
What to take from the list
- Research, safety and computing capacity each have a cofounder at the top, so none is an afterthought in the org chart.
- The company's leadership is stable at the cofounder level, though roles have shifted, as McCandlish's move to chief architect shows.
- The profile is light on new facts. It is a quick reference, not an investigation.
That last point is worth keeping in mind. The piece does not reveal strategy or tell you anything about pricing, and it is best read as an introduction. If you want to judge the company, its products and its published decisions will tell you more than a list of biographies.
What to watch
- Whether Anthropic's leadership roles change again as the company grows and goes public.
- How cofounders such as Jack Clark and Chris Olah shape public discussion of AI's social effects and interpretability.
Try it
Pick one cofounder's area, such as interpretability or scaling laws, and search for how that work is described in Anthropic's own published research.
Sources and links
First reported by Open Magazine. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.