Key Takeaways
- The 21 APEC economies issued the “Chengdu statement” on AI, emphasizing security, data protection, intellectual‑property rights, and support for open‑source development with strong security assurances.
- The statement signals a shift from the libertarian origins of open‑source AI toward greater state oversight, especially as China leads recent open‑source model releases.
- Experts note that the language “strong security assurance” allows security‑conscious economies to back open models while demanding testing, transparency, and deployment controls.
- Discussions at the Chengdu meeting also highlighted the need to prepare for quantum‑computing era challenges, particularly trusted encryption for financial systems.
- While the agreement shows multilateral consensus on trusted open‑source AI, the rapid pace of technological change may outstrip policy responses.
Overview of the Chengdu Statement
On July 23, 2026, Li Lecheng, China’s Minister of Industry and Information Technology, chaired a press conference in Chengdu as part of APEC Digital Weeks. The 21 member economies—including the United States and China—released the “Chengdu statement” on artificial intelligence. The document stressed respect for “security, data protection and intellectual property rights,” while calling for support of open‑source development that incorporates “strong security assurance through development and deployment.” Li Lecheng highlighted that the statement marked a ministerial‑level first: “The APEC AI statement is the first one to include open-source cooperation at a minister level,” he said, underscoring the significance of the collective endorsement.
Open‑Source AI and Growing State Oversight
The Chengdu statement reflects a broader trend where open‑source AI is moving away from its libertarian roots toward frameworks that involve more governmental supervision. By explicitly aligning around “open-source ecosystems and compute infrastructure,” the APEC economies signalled a shift toward open‑weight systems that are integrated with state‑coordinated energy, telecom, and digital networks. Winston Ma, adjunct professor of law at New York University, observed that this alignment confirms “Asia‑Pacific is moving away from closed models to open-weight systems that are combined with state-coordinated energy, telecom and digital infrastructure.” The emphasis on security assurances suggests that governments wish to reap the benefits of open collaboration while retaining oversight mechanisms.
China’s Leadership and U.S. Allegations
China has been at the forefront of recent open‑source AI advances, releasing models such as DeepSeek and GLM 5.2 that are freely available for download and use. In contrast, many U.S. firms like Anthropic continue to offer only closed, pay‑to‑use models. During the Chengdu press conference, Li Lecheng acknowledged the competitive environment, noting that “the U.S. has claimed that China steals from American tech to advance its own open-source capabilities.” While he did not elaborate on the allegations, his remarks indicated awareness of the sensitivities surrounding intellectual‑property claims and the need for transparent dialogue to mitigate mistrust.
Security, Trust, and Deployment Controls
A pivotal phrase in the Chengdu statement—“strong security assurance”—was highlighted by Wei Sun, principal analyst for artificial intelligence at Counterpoint Research, as a way to give “more security‑conscious economies room to support open models while still demanding testing, transparency, data protection and deployment controls.” She added that the debate is evolving “beyond open vs. closed and towards: who can now build an open ecosystem that is also trusted enough for governments and enterprises to deploy.” This nuanced stance allows nations to weigh the advantages of open‑source innovation against the imperative of safeguarding critical infrastructure and data.
Expert Perspectives on Trusted Open‑Source AI
The Chengdu gathering produced a range of viewpoints on how to operationalize trusted open‑source AI. Li Lecheng stressed that all participants “recognized the risks of AI, and encouraged dialogue across the government, private sector and academia to share information about cybersecurity, supply chain resilience and online scams.” Such cross‑sector communication aims to build a shared understanding of threats and best practices. Janet De Silva, chair of the APEC Business Advisory Council Digital and Innovation Working Group, urged ministers to look beyond current AI challenges and prepare for the quantum‑computing era: “We need trusted encryption to ensure the security of our entire financial system, ensuring that quantum computing is not only a security measure but also an opportunity to bring competitiveness.” Her comment underscored the forward‑looking nature of the discussions, linking AI governance to emerging cryptographic needs.
Quantum Computing and Future‑Proofing
While the Chengdu statement primarily addressed AI, the ancillary conversation about quantum computing revealed concerns that rapid advances in quantum capabilities could undermine existing encryption standards. Delegates warned that without proactive measures, the financial system—and by extension, broader economic stability—could be exposed to new vulnerabilities. The call for “trusted encryption” reflects a consensus that preparing for quantum threats must begin now, integrating quantum‑resistant cryptographic techniques into the open‑source AI ecosystems that the APEC economies aim to nurture.
Conclusion: Balancing Innovation and Control
The Chengdu statement represents a notable step toward multilateral agreement on fostering open‑source AI under a framework of security assurances and state collaboration. By bringing together diverse economies—including strategic competitors like the U.S. and China—the APEC forum has opened a channel for sharing best practices on cybersecurity, supply‑chain resilience, and fraud prevention. However, as experts warn, the velocity of technological change, especially in quantum computing, may outpace the policy mechanisms established today. Continued dialogue, transparent security testing, and adaptive regulatory approaches will be essential if the envisioned trusted open‑source ecosystem is to deliver both innovation and protection for governments and enterprises alike.
https://www.cnbc.com/2026/07/24/china-ai-open-source-apec.html