What is the Chengdu Statement?

It’s a joint declaration adopted by all 21 APEC economies on July 23, 2026, in Chengdu, China, backing open-source AI development paired with strong security safeguards, alongside broader digital-cooperation goals.

Is the Chengdu Statement legally binding?

No. Like other APEC ministerial statements, it reflects a shared policy direction rather than enforceable law — APEC operates by consensus and has no authority to require members to adopt specific rules.

Which economies signed the statement?

All 21 APEC members endorsed it, including the United States, China, Japan, Russia and South Korea, under the chairmanship of China’s Minister of Industry and Information Technology, Li Lecheng.

How does this square with the US-China dispute over Moonshot AI's Kimi K3?

The two governments still disagree over specific allegations — Washington accuses Moonshot AI of improperly distilling Anthropic’s model — but the Chengdu Statement shows they can still agree on baseline principles for open-source AI and security even amid that rivalry.

US, China Among 21 APEC Nations Backing Open-Source AI

Ministers from all 21 Asia-Pacific Economic Cooperation (APEC) economies, including the United States and China, adopted a joint statement in Chengdu, China, on July 23 backing open-source artificial intelligence paired with stronger security safeguards.

The Chengdu Statement, titled “Digital Technologies and AI for the Empowerment of an Asia-Pacific Community,” is the first APEC ministerial-level document to explicitly endorse open-source AI, according to the text published by APEC. It calls on member economies “to support open-source models and projects that employ strong security assurance through development and deployment,” while urging respect for intellectual property and data protection.

China’s Minister of Industry and Information Technology, Li Lecheng, chaired the meeting and said the 21 economies reached “consensus… on promoting the digital and intelligent transformation of traditional industries,” pledging to keep using the APEC platform to carry out the agreement. The bloc’s next ministerial meeting is set for Vietnam in 2027.

Cooperation despite rivalry

The statement also voiced “deep concern over the escalation of online scam activities” and committed members to sharing information on cybersecurity, data protection and supply-chain resilience.

The agreement lands weeks after Washington accused Moonshot AI of illicitly distilling Anthropic’s Fable model to build its Kimi K3 system — a dispute that has sharpened US-China tension over AI intellectual property, even as the two governments prepare for separate talks planned for September. Despite that friction, both were signatories in Chengdu, suggesting the rivals can still find common ground on baseline AI-governance principles while competing commercially. Wei Sun, an analyst at Counterpoint Research, said the consensus signals debate shifting “beyond open vs. closed” toward building open-source ecosystems that governments can deploy with confidence.

The statement also sets out priorities for digital infrastructure, connectivity and AI-literacy programs, and calls for wider AI use in agriculture, healthcare, education and transport, building on APEC’s 2026-2030 AI Initiative.

Cooperation despite rivalry

Frequently asked questions