China’s Minister of Industry and Information Technology Li Lecheng spoke at a press conference in Chengdu, China on July 23, 2026, as part of APEC Digital Week.
Evelyn Chen | CNBC
CHENGDU, China — Governments are increasingly seeking control over artificial intelligence as companies release more powerful open source models, a new multilateral statement shows.
Twenty-one APEC member countries, including the United States and China, released the Chengdu Statement on AI, emphasizing respect for “security, data protection, and intellectual property rights” and calling for support for open source development.
The statement released Thursday also specifies support for models and projects that “employ strong security guarantees throughout development and deployment.”
These details reflect how open source is moving away from its libertarian roots and toward more involvement of state surveillance. China has recently led the way in open source AI development, despite claims by the United States that the technology is stolen from American technology.
Open source models from Chinese companies such as DeepSeek and GLM 5.2 are free to use and download. In contrast, US companies such as Anthropic only offer closed, pay-as-you-go AI models.
Winston Ma, an adjunct professor at New York University School of Law, said the Chengdu statement’s “explicit collaboration around open source ecosystems and computing infrastructure” confirms that the Asia-Pacific region is moving from a closed model to an open weight system combined with state-coordinated energy, communications and digital infrastructure.
APEC’s AI statement is the first to include ministerial-level open source cooperation, according to China’s Industry and Information Technology Minister Li Lecheng, who chaired Thursday’s summit.
He said all parties are aware of the risks of AI and encourages dialogue across government, the private sector and academia to share information on cybersecurity, supply chain resilience and online fraud.
The comments came amid growing reports that the U.S. and Chinese governments are seeking greater control over domestic AI capabilities.
“It makes sense to have 21 countries, including both China and the United States, recognize trustworthy open source AI as something worth supporting,” said Wei Sun, principal analyst for artificial intelligence at Counterpoint Research.
“The term ‘strong security guarantees’ gives more security-conscious economies room to support open models while requiring testing, transparency, data protection, and deployment control,” he said. “The debate is moving beyond open versus closed to who can build an open ecosystem that is trustworthy enough for governments and businesses to deploy.”
But technology could eventually advance much faster.
Earlier on Thursday, APEC Business Advisory Council Digital Innovation Working Group Chair Janet de Silva urged ministers to prepare for the era of quantum computing.
“Ensuring the security of the entire financial system requires reliable encryption. This makes quantum computing not only a security measure, but also an opportunity for competitive advantage.”
