From stuffed animals to Anthropic’s presidency: Daniela Amodei’s journey in artificial intelligence

A stuffed animal council for resolving corporate conflicts

Daniela Amodei, president of Anthropic, once relied on a group of stuffed animals to reflect on management dilemmas. The cat Jinji suggested canceling meetings, the pink bear Maura demanded attention to detail, and the panda Beary Bonds represented generosity. According to The Atlantic, she and her husband, Holden Karnofsky, attributed personalities to the animals to simulate leadership scenarios, even during their time at OpenAI. The practice served as an exercise in personality analysis, though few colleagues knew about it.

From literature to technology: a multidisciplinary career

Daniela graduated with a literature degree from the University of California, Santa Cruz in 2009, with additional studies in politics. She began her career in international development, working for NGOs in Washington and on public health projects in East Africa. She also worked on an election campaign and in the U.S. Congress. In 2022, speaking on the Future of Life Institute podcast, she said she was seeking greater impact, which led her to Stripe. There, she witnessed the growth of a payment startup and realized the power of a small team to create technology with economic reach.

At Stripe, her role evolved from recruitment to risk operations, where she observed the contrast between rule-governed systems and AI labs still defining security practices. This experience prepared her for the vice presidency of security and policy at OpenAI, where she combined people management and operational controls.

Founding Anthropic: from the backyard to the laboratory

In December 2020, Daniela left OpenAI alongside her brother Dario and other former colleagues to found Anthropic, the company behind the Claude language model. Launched in March 2023, Claude competes with ChatGPT. The first meetings took place outdoors, with masks and social distancing due to pandemic restrictions. The goal was to create an organization dedicated to research and safety, but it required structuring teams, raising funds, and developing a language model from scratch.

After testing with Notion, Quora, and DuckDuckGo, Claude began offering writing, document summarization, and programming features. The team grew from about 60 to approximately 140 people during development. In a July 2023 interview with Stripe, Daniela emphasized that security needs to be attractive to customers, as consumers and businesses do not want to pay for systems that generate false information or harmful content. “The market seeks trust and security,” she said, acknowledging the potential conflicts between ethical principles and commercial interests.

Between the laboratory and the business: leadership and routine

Daniela currently chairs Anthropic and its board of directors, managing the executive team and coordinating research, engineering, product, governance, and sales. Dario, as CEO, leads research direction and strategic vision. Her routine includes two hours at the start of the day for reflection and larger projects, before meetings and strategic decisions. She also strives to meet new employees and understand their experiences, a task that becomes more challenging as the company grows.

In a June 2026 conversation published by Stanford, Daniela described herself as a generalist and reaffirmed that she would choose the same academic path again, valuing curiosity and interdisciplinarity. In an interview with ABC News, cited by the University of California, Santa Cruz, she stated that “studying the humanities will be more important than ever.”

From the stuffed animal council to the presidency of one of the leading AI labs, Daniela Amodei continues to apply her ability to understand different personalities, now teaching machines how to converse.