Mark Zuckerberg says strengthening U.S. competitiveness should come before regulating China’s AI

[Financial News, New York = Lee Byung-chul] Mark Zuckerberg, CEO of Meta Platforms, said that the U.S. government's move to regulate Chinese artificial intelligence (AI) models is "not an effective solution." He also warned against so-called "regulatory capture," in which U.S. AI companies exert excessive influence over the regulatory process.
In an interview with the Financial Times (FT) on the 29th local time, Zuckerberg said, "Banning cutting-edge Chinese AI in the United States is not an effective solution."
He said the United States should focus on systematically removing bottlenecks and obstacles in the industry rather than blocking Chinese models if it wants to maintain AI competitiveness. Zuckerberg explained that because the AI race is important to national security and economic prosperity, policies should help U.S. companies become more competitive.
His remarks came as the Trump administration reviews regulations on Chinese AI companies. The U.S. government is investigating allegations that Chinese AI startup Kimi K3 used a "distillation" technique to train on U.S. AI models. Treasury Secretary Scott Bessent also mentioned last week the possibility of sanctions against Chinese AI labs if intellectual property rights violations are confirmed.
Zuckerberg also presented a vision called "AI for Everyone" and reaffirmed his support for open AI models. He said open models, which users can modify directly and run on their own hardware, are important for the development of the AI ecosystem.
He also voiced concern that AI technology could become concentrated in the hands of a few companies. "If companies with vested interests review one another, regulatory capture can become a problem," he said. "When asked whether leading AI labs want open-source models to succeed, there were mixed signals."
On the Trump administration's push to introduce a voluntary safety verification system before AI models are released, he said peer review and vetting themselves are positive steps. Still, he stressed that leading AI companies should not have excessive influence over the verification process.
Zuckerberg also addressed the debate over AI safety, saying, "Restricting the most powerful AI is not the answer to cybersecurity problems." He added, "Advanced AI can also be used to identify vulnerabilities and fix them."




[email protected] Lee Byung-chul Reporter