Chinese AI researcher leaves Anthropic over its 'anti-China' stance

(ECNS) -- A prominent Chinese artificial intelligence researcher has left U.S.-based AI firm Anthropic to join a rival company, citing what he called Anthropic's "anti-China statements" as the primary reason for his departure.

Yao Shunyu left Anthropic after less than a year and joined Google DeepMind because he strongly opposed the start-up's description of China as an "adversarial nation" and its broader rhetoric, according to a post on his website Monday.

Last month, Anthropic designated China as an "adversarial nation" in an internal file when it announced its decision to prohibit subsidiaries of Chinese companies worldwide from accessing its services.

"Although, to be clear, I believe most of the people at Anthropic disagree with such a characterization, I don't think there is a way for me to stay," Yao wrote.

Yao graduated from Tsinghua University and earned a PhD in theoretical and mathematical physics from Stanford University.

He joined Anthropic in October 2024, working on the development of its Claude 3.7 Sonnet model, which was released in February this year.

Yao chose Anthropic because it was one of the best places for physicists to start their careers in AI research, he wrote in his post.

"Relative to physics, AI moves insanely fast and, looking back, I am surprised by how much has happened in the past year," he wrote.

According to the South China Morning Post, several U.S. AI companies, including OpenAI, have intensified their negative rhetoric toward China, with some directly naming Chinese competitors like DeepSeek.

An anonymous former employee revealed that some researchers from China and other countries within OpenAI have expressed unease over the company's related statements.

(By Zhang Dongfang)