APEC Ministers Agree on Framework for Governing Open-Source AI, Emphasizing Security and Oversight

CHENGDU, China – In a significant multilateral development, the 21 member economies of the Asia-Pacific Economic Cooperation (APEC) have issued a joint statement, dubbed the "Chengdu statement on AI," signaling a growing international consensus on the need for governmental oversight of artificial intelligence, particularly as powerful open-source models proliferate. The statement, released on Thursday, July 23, 2026, underscores a shift in the landscape of AI development and governance, moving towards a framework that balances the benefits of open-source innovation with crucial considerations for security, data protection, and intellectual property rights.

China’s Minister of Industry and Information Technology, Li Lecheng, who chaired the ministerial meeting, highlighted the groundbreaking nature of the statement, noting it as the first to include open-source cooperation at a ministerial level. Speaking at a press conference in Chengdu as part of APEC Digital Weeks, Minister Li emphasized that all participating economies recognized the inherent risks associated with artificial intelligence. He further stated that the dialogue encouraged collaboration across government, the private sector, and academia to facilitate the sharing of vital information concerning cybersecurity, supply chain resilience, and the mitigation of online scams.

The "Chengdu statement" explicitly calls for the support of AI models and projects that "employ strong security assurance through development and deployment." This detail is seen by experts as a reflection of how open-source principles are evolving, incorporating a greater degree of state oversight beyond their traditionally libertarian roots. This development is particularly noteworthy given China’s recent advancements in open-source AI, a sector where the United States has raised concerns about alleged technology theft.

Companies like China’s DeepSeek and GLM 5.2 are offering their advanced AI models for free use and download, positioning them as open-source alternatives to closed, pay-to-use models offered by some U.S. companies, such as Anthropic. This divergence in approach highlights a broader global trend. Winston Ma, an adjunct professor of law at New York University, commented on the statement’s significance, stating that the "explicit alignment around open-source ecosystems and compute infrastructure" indicates that the Asia-Pacific region is increasingly favoring open-weight systems. These systems, he noted, are being integrated with state-coordinated energy, telecommunications, and digital infrastructure.

The impetus for such a multilateral agreement stems from mounting reports suggesting that both Washington and Beijing are seeking to implement tighter controls over their domestic AI capabilities. Wei Sun, principal analyst for artificial intelligence at Counterpoint Research, described the APEC statement as "meaningful," particularly in achieving recognition for trusted open-source AI among 21 economies, including both China and the United States. She elaborated that the inclusion of the phrase "strong security assurance" provides more security-conscious nations with the latitude to support open models while simultaneously demanding rigorous testing, transparency, data protection, and deployment controls. Sun posits that the global discourse surrounding AI is moving beyond a binary choice between "open vs. closed" and is now focusing on "who can build an open ecosystem that is also trusted enough for governments and enterprises to deploy."

However, the pace of technological advancement poses a continuous challenge to regulatory efforts. Earlier on Thursday, Janet De Silva, chair of the APEC Business Advisory Council’s Digital and Innovation working group, urged ministers to proactively prepare for the advent of the quantum computing era. She stressed the critical need for "trusted encryption" to safeguard the integrity of the global financial system, asserting that quantum computing presents not only a significant security concern but also a potential avenue for enhanced competitiveness.

The APEC Digital Weeks initiative, of which Minister Li’s address was a part, serves as a crucial platform for fostering dialogue and collaboration among member economies on critical digital economy issues. The inclusion of AI governance, with a specific focus on the evolving nature of open-source development, underscores the region’s commitment to navigating the complex challenges and opportunities presented by this transformative technology. The "Chengdu statement" represents a foundational step in establishing international norms and cooperation frameworks for AI, aiming to foster innovation while ensuring responsible development and deployment that aligns with global security and ethical considerations.

The emphasis on "security, data protection and intellectual property rights" within the statement reflects a growing awareness of the potential vulnerabilities and risks associated with advanced AI technologies. As open-source models become more powerful and accessible, the need for robust safeguards and collaborative governance mechanisms becomes paramount. The statement’s call for supporting models that exhibit "strong security assurance" signals a desire to promote AI development that is not only innovative but also inherently secure and trustworthy.

The participation of a diverse group of economies, ranging from major technological powers like the United States and China to smaller developing nations, in reaching this consensus highlights the global reach and impact of AI. The collaborative approach adopted by APEC in addressing these complex issues aims to create a more stable and predictable environment for AI innovation and adoption across the Asia-Pacific region and beyond. The ongoing dialogue and the commitment to information sharing among governments, industry, and academia are crucial for staying ahead of emerging threats and ensuring that AI technologies are developed and utilized for the benefit of all. The inclusion of quantum computing preparedness further demonstrates a forward-looking perspective, acknowledging that the technological landscape is constantly evolving and requires continuous adaptation and strategic planning.