U.S. and Allies Endorse Open‑Source AI with Robust Security Framework at Chengdu Summit

Li Lecheng, China’s Minister of Industry and Information Technology, delivered remarks at a press conference on July 23, 2026, in Chengdu, during the APEC Digital Weeks.

Evelyn Cheng | CNBC

CHENGDU, China — As companies roll out increasingly powerful open‑source AI models, governments are intensifying efforts to regulate artificial intelligence, according to a new multilateral statement.

The 21 APEC member economies — including the United States and China — issued the Chengdu Statement on AI, stressing respect for security, data protection, and intellectual property rights while advocating for the advancement of open‑source development.

The statement, released on Thursday, also pledged support for models and projects that incorporate robust security assurance throughout development and deployment.

These details illustrate a shift of open‑source development away from its libertarian origins toward greater state oversight, with China leading recent open‑source AI initiatives amid U.S. allegations of technology theft.

Open‑source models from Chinese firms such as DeepSeek and GLM 5.2 are freely available for use and download, unlike many U.S. companies — including Anthropic — that provide only closed, subscription‑based AI solutions.

The Chengdu Statement’s explicit focus on open‑source ecosystems and coordinated compute infrastructure signals Asia‑Pacific’s transition from closed models to open‑weight systems integrated with state‑coordinated energy, telecommunications, and digital infrastructure, according to Winston Ma, an adjunct professor of law at New York University.

The APEC AI statement marks the first minister‑level inclusion of open‑source cooperation, Li Lecheng, China’s minister of industry and information technology, said while chairing the leaders’ meeting on Thursday.

He noted that all parties acknowledged AI risks and called for dialogue among governments, the private sector, and academia to share information on cybersecurity, supply‑chain resilience, and online scams.

These remarks follow growing reports that Washington and Beijing are pursuing tighter controls over domestic AI capabilities.

“Securing the endorsement of 21 economies — including both China and the United States — for trusted open‑source AI is meaningful,” said Wei Sun, principal AI analyst at Counterpoint Research.

“The phrase ‘strong security assurance’ enables security‑focused economies to back open models while still requiring rigorous testing, transparency, data protection, and deployment controls,” she explained. “The discussion has evolved beyond the simple open‑vs‑closed divide, focusing instead on who can construct a trusted open ecosystem suitable for government and enterprise deployment.”

Nevertheless, technological progress may accelerate.

Janet De Silva, chair of the APEC Business Advisory Council’s Digital and Innovation Working Group, urged ministers on Thursday to prepare for the quantum computing era.

“We need reliable encryption to safeguard our entire financial system, ensuring that quantum computing serves not only as a security measure but also as a catalyst for competitiveness.”

Source link