Kimi K3's CEO said Trump had blocked his U.S. studies

Signals Inbox·July 19, 2026·Frontier AI

Kimi K3 has turned its founder Yang Zhilin into a symbol of America’s struggle to retain foreign AI talent. But the viral claim that Donald Trump blocked Yang from studying in the U.S. does not fit the record. Yang completed a Carnegie Mellon PhD in 2019, worked at Google Brain and Meta, and then chose to build companies in China.

Q1What is the official record?

Yang’s Carnegie Mellon thesis record shows that he completed his U.S. PhD in 2019. He had already spent four years studying there, interned at Google Brain and Meta, and worked with leading American researchers. The post’s claim that Trump prevented him from studying in America is therefore misleading.

Q2Why did Yang return to China?

His former PhD supervisor says Yang wanted to start his own company and believed he would regret not trying. After graduating, Yang returned to China, helped build major Chinese AI projects, and eventually founded Moonshot AI in 2023. There is no solid evidence that Trump personally forced that decision.

Q3Why is this story exploding now?

Because Kimi K3 made Yang strategically important. The new model is competing with leading American systems in coding and agent tasks while offering much lower prices. A founder who once studied in Pittsburgh is now building one of China’s strongest challenges to U.S. frontier AI.

Q4Does America still have a talent problem?

Yes, just not in the simple way the tweet claims. The U.S. still attracts many of the world’s best researchers through universities and technology companies. The harder question is whether they can easily stay, start companies, and plan their lives without years of visa uncertainty. China can benefit whenever that path feels easier elsewhere.

Q5What changed in the last year?

Washington increased scrutiny of Chinese students, paused some visa interviews, and announced plans to revoke visas connected to sensitive fields or the Chinese Communist Party. At the same time, Chinese labs released increasingly competitive open models. That combination makes every successful Chinese founder with a U.S. education look like evidence in a much bigger political fight.

Q6So what is the real signal?

Not that Trump blocked Yang’s education. The real signal is that America can train exceptional AI researchers without keeping the companies they later create. Kimi K3 turns that abstract concern into a visible example: U.S. universities helped develop the talent, while China captured the company, the model, and the strategic upside.