China pushes for AI safety as G7 summit wraps up without Beijing

0
2
Undersecretary of Defense: Anthropic's actions show it was a

The U.S. and China flags are seen on the day of a bilateral meeting between the U.S. and China on May 10, 2025 in Geneva, Switzerland.

Keystone/eda/martial Trezzini | Via Reuters

BEIJING – Senior Chinese officials on Wednesday emphasized Beijing’s plans to share artificial intelligence globally and securely, the latest sign of how the United States and China are promoting different approaches to the technology.

“China is accelerating the construction of a global AI cooperation organization and welcomes all parties to join,” Wang Yi, China’s top diplomat, told reporters in Mandarin Chinese, according to a CNBC translation. He emphasized that technology should serve people’s needs.

Wang spoke at the release of China’s white paper on global governance, which criticized trade wars and emphasized support for the global south. The category broadly refers to less developed economies, particularly countries outside the United States and Europe.

Wang’s comments came as the US stepped up efforts to restrict foreign access to leading US-developed AI models.

During a summit in France this week, wealthy Group of Seven countries – the United States, Britain, France, Germany, Canada, Italy and Japan – discussed a plan to give “trusted partners” access to the U.S. AI models, Reuters reported on Tuesday, citing three diplomatic sources. CNBC could not independently confirm the report and has reached out to the White House for comment.

U.S. AI models are also typically available only by subscription, while China’s efforts have focused on cheap or free AI models that are often fully downloadable.

Zhao Haibing, vice chairman of China’s top economic authority, spoke alongside Wang on Wednesday and rejected “closed, exclusive and monopolistic approaches to technology development.”

Instead, Zhao highlighted China’s efforts to deepen international AI cooperation through the BRIC nations and the Shanghai Cooperation Organization, an annual meeting of countries such as Russia and Iran that initially focused on security.

Zhao also noted China’s AI Capacity Building for All initiative, the United Nations’ support in leading global AI governance, and efforts to help developing countries with technology and talent.

The U.S. and China separately said last month that both sides would work on AI guardrails, but few details were revealed.

Beijing has made far-reaching proposals for global cooperation in the last twelve months.

Chinese President Xi Jinping proposed the Global Governance Initiative at a Chinese-hosted SCO meeting late last summer.

A few weeks earlier, Chinese Premier Li Qiang announced at an annual AI conference in Shanghai that the Chinese government had proposed establishing a global AI cooperative organization. Li’s comments came just days after the Trump administration announced an AI action plan that included supporting U.S. technology development abroad.

Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.