Full Breakdown
China’s AI Strategy: Covert Influence, Domestic Censorship, and Shifting Trade Policies
7/10/2026, 11:57:59 AM
Background & Context
The Chinese Communist Party has long turned new technologies into tools for surveillance and propaganda. A 2022 mandate required AI algorithms to uphold “mainstream values” and “actively spread positive energy,” effectively binding domestic models to party lines. The recent release of the open-source GLM-5.2 model, which rivals leading U.S. systems, has intensified U.S. concerns about Chinese AI competitiveness.
Covert Influence Operations
OpenAI disclosed a campaign that used ChatGPT to generate an English-language comic portraying AI data-center costs as prohibitive, aiming to slow U.S. AI infrastructure build-out. The accounts involved were linked to a private technology firm serving provincial officials. A separate effort sought to discredit Japanese Prime Minister Sanae Takaichi, employing “hundreds of staff” and “thousands of fake accounts across scores of platforms.” Researchers note that AI now enables faster, more granular censorship inside China, with state-sponsored sources biasing chatbot answers in Chinese.
Policy Shifts on AI Hardware
Chinese officials have granted top AI firms—Alibaba Group Holding Ltd., ByteDance Ltd., and DeepSeek—permission to purchase a limited allotment of Nvidia H200 chips, the most powerful AI processors available until Nvidia’s next-generation Blackwell line. Companies must justify the quantity and purpose of each request; sources suggest the total may stay below 200,000 units, far short of the >400,000 chips a single data center would require.
Official Statements & Responses
- “China opposes groundless smear and ill-intentioned association” and seeks to “ensure AI is a force for good and for all.” — Chinese embassy spokesperson, Washington, D.C.
- “We advocate that China and the United States achieve mutual benefit and win-win outcomes through cooperation, and we oppose the politicization, instrumentalization, and weaponization of technological and economic issues.” — Liu Chang, Chinese embassy spokesperson
- Nvidia CFO Colette Kress said the company is “uncertain whether any imports will be allowed into the country,” reflecting ongoing U.S. export controls.
Criticism & Opposition
Republican lawmakers and U.S. officials have linked the data-center smear effort to a broader foreign-influence campaign that threatens domestic AI investment. Analysts warn that Chinese-origin models such as GLM-5.2 could “hoover up sensitive data” and pose national-security risks, potentially prompting future regulation.
Data & Statistics
- OpenAI’s probe identified “at least hundreds of staff” and “thousands of fake accounts across scores of platforms” in the Japan disinformation operation.
- The Australian Strategic Policy Institute found AI-driven censorship now “far more pervasive, granular and reflexive” than a few years ago.
- Chinese AI models accounted for nearly half of all open-source AI downloads from February 2025 to 2026, according to Hugging Face data.
Conflicting Reports & Gaps
OpenAI reports the data-center comic campaign “gained little traction,” while some U.S. officials argue it has contributed to broader opposition to AI infrastructure. The exact quota for H200 chip sales remains unspecified; sources cite “less than 200,000” but no official figure has been released.
Verbatim Quotes
- “China opposes groundless smear and ill-intentioned association” — Chinese embassy spokesperson, Washington, D.C.
- “What AI brings to the game is, it helps plan information campaigns and it helps to execute them,” — Kenton Thibaut, senior fellow, Atlantic Council
- “We advocate that China and the United States achieve mutual benefit and win-win outcomes through cooperation, and we oppose the politicization, instrumentalization, and weaponization of technological and economic issues,” — Liu Chang, Chinese embassy spokesperson
- “We are uncertain whether any imports will be allowed into the country,” — Colette Kress, Nvidia CFO
What’s Next
Chinese authorities are expected to finalize the H200 chip purchase quota in the coming weeks, while U.S. regulators continue to assess export-control policies for advanced AI hardware.
