Anthropic CEO Dario Amodei Calls for Independent AI Oversight, Drawing on His Scientific and Industry Background
Dario Amodei, CEO and co-founder of Anthropic, has publicly proposed that frontier AI companies accept "embedded evaluators" — independent third-party teams with ongoing access to AI development processes. He cited METR, an AI safety testing laboratory, as an example of the kind of organization that could fill that role. Amodei holds a Ph.D. in biophysics from Princeton and previously served as vice president of research at OpenAI before leaving in 2020 to co-found Anthropic.
Dario Amodei's path to leading one of the AI industry's most prominent safety-focused companies ran through academia, Silicon Valley, and a series of escalating concerns about how the technology he was helping build would be governed. He earned his Ph.D. in biophysics from Princeton in 2011 and later became a postdoctoral scholar at the Stanford University School of Medicine before shifting into industry roles at Baidu, Google Brain, and ultimately OpenAI.
At Baidu, Amodei worked on speech recognition through machine learning. At Google, he began working on safety while contributing to the company's neural-network research. At OpenAI, he rose to vice president of research as the company developed its GPT-2 and GPT-3 models. He departed in 2020, saying in a Bloomberg interview that he had lost confidence in OpenAI's commitment to safety over financial interests: "When you feel that you can't trust someone, when you feel that their values are not what they say they are… that makes it very hard to continue to work with a company," Fox News reported him saying.
After leaving, Amodei co-founded Anthropic, which publishes a guiding document — described as a kind of constitution — for its flagship AI, Claude. The company has declined work it determined would enable autonomous battlefield targeting or mass surveillance, a stance that Fox News reported cost Anthropic a $200 million Pentagon contract.