As tensions between Washington and Beijing grow over allegations that Moonshot AI’s Kimi K3 distilled capabilities from Anthropic’s Claude Fable 5, the world’s two biggest AI powers have found rare common ground on one issue: the future of open-source AI.
At the Asia-Pacific Economic Cooperation (APEC) Digital and AI Ministerial Meeting in Chengdu, China, the United States, China and 19 other member economies adopted a joint statement supporting open-source AI while calling for stronger protections around security, data privacy and intellectual property.
The agreement comes at a pivotal moment for artificial intelligence. Governments are racing to shape AI policy as open models become more capable and more widely adopted across businesses, research institutions and public agencies. At the same time, concerns over model safety, cybersecurity and intellectual property are pushing policymakers to rethink how open AI should be governed.
21 APEC economies, including the US, endorse open-source AI with security, data protection and IP safeguards
Released Thursday, the Chengdu Statement calls for supporting “the development of open-source models and projects” while encouraging systems that “employ strong security assurance through development and deployment.” The language reflects a growing consensus that openness and security no longer need to be treated as competing priorities.
The statement goes beyond endorsing open-source AI. It positions trusted AI, digital infrastructure and regional cooperation as interconnected priorities, highlighting investments in resilient digital ecosystems, secure connectivity and cross-border collaboration. The approach suggests that APEC economies increasingly view AI as critical infrastructure that underpins economic growth, innovation and public services, rather than simply another software technology.
That marks a notable shift from the early philosophy of open-source software, which emphasized openness with minimal oversight. Today’s governments appear to be moving toward a different model: encouraging broader access to AI while pairing it with stronger security standards, transparency and governance requirements.
The broader backdrop remains one of intensifying AI competition.
China has emerged as one of the leading forces behind open-source AI development through models such as DeepSeek and GLM, while many leading U.S. AI companies, including Anthropic, continue to focus on proprietary commercial models. Meanwhile, Washington has tightened restrictions on advanced AI chips and other technologies, citing national security concerns, and Chinese AI developers have faced growing scrutiny following allegations that some models were trained using outputs from leading American AI systems.
Yet the APEC agreement shows that competition and cooperation can exist at the same time. Even as governments compete for leadership in artificial intelligence, they continue to search for common principles for deploying the technology safely and responsibly.
Winston Ma, adjunct professor of law at New York University, said the Chengdu Statement’s “explicit alignment around open-source ecosystems and compute infrastructure” confirms that Asia-Pacific is moving away from closed AI models toward open-weight systems supported by coordinated investments in energy, telecommunications and digital infrastructure.
China’s Minister of Industry and Information Technology, Li Lecheng, who chaired Thursday’s meeting, said the agreement represents the first APEC ministerial statement to include cooperation on open-source AI, CNBC reported. He said member economies recognized both the opportunities and risks of artificial intelligence and agreed to strengthen collaboration among governments, businesses and academia on cybersecurity, supply chain resilience and online fraud.
Open-source AI gains global support as APEC calls for stronger security protections
Industry analysts view the agreement as a sign that the debate around AI governance is entering a new phase.
“Getting 21 economies, including both China and the United States, to recognize trusted open-source AI as something worth supporting is meaningful,” said Wei Sun, principal analyst for artificial intelligence at Counterpoint Research.
“The phrase ‘strong security assurance’ gives more security-conscious economies room to support open models while still demanding testing, transparency, data protection and deployment controls,” she said. “The debate is moving beyond open vs. closed and towards: who can now build an open ecosystem that is also trusted enough for governments and enterprises to deploy.”
That observation mirrors the broader direction outlined in the Chengdu Statement. Rather than choosing between open and closed AI, APEC’s ministers appear to be endorsing a third path: trusted open-source AI backed by stronger security standards, resilient digital infrastructure and regional cooperation.
The discussion also extended beyond today’s AI systems. Janet De Silva, chair of the APEC Business Advisory Council’s Digital and Innovation Working Group, urged ministers to begin preparing for the quantum computing era, calling for trusted encryption to protect financial systems while ensuring quantum technologies strengthen both security and economic competitiveness.
For now, the Chengdu Statement offers a rare reminder that geopolitical rivals can still cooperate on issues with global consequences. While Washington and Beijing continue to compete for AI leadership, both sides joined 19 other Asia-Pacific economies in signaling that the future of artificial intelligence may depend not on choosing between open and closed models, but on building open AI ecosystems that governments, businesses and citizens can trust.
