https://www.politico.com/news/2026/06/10/openai-china-ai-data-centers-report-00957612
OpenAI reports that China-linked operatives have utilized ChatGPT to infiltrate and manipulate American public discourse on key issues, including trade tariffs and U.S. domestic AI infrastructure. Furthermore, U.S. authorities and OpenAI have documented Chinese espionage campaigns attempting to systematically "distill" (copy) proprietary technology from American AI companies. [1, 2, 3, 4]
Key details regarding these intelligence and infiltration operations include:
1. Covert Influence Operations
- Targeting AI Data Centers: OpenAI identified a suspected Chinese campaign designed to stir up local opposition to U.S. AI data centers. By posing as Americans, operatives attempted to exploit existing domestic debates over energy grids to undermine U.S. artificial intelligence expansion. [1, 2, 3]
- Fueling Anti-Tariff Rhetoric: A second operation used ChatGPT to generate political cartoons and social media comments criticizing the Trump administration's tariffs and framing U.S. tech policies negatively. [1]
- Impact: According to OpenAI, these operations were small-scale and failed to gain meaningful organic traction or sway authentic public opinion. [1, 2]
2. Intellectual Property and Technology Theft
- Model Distillation: OpenAI and the U.S. White House have warned lawmakers that Chinese firms (such as DeepSeek) are using illicit methods to "distill" outputs from top-tier American AI models. This allows foreign actors to replicate breakthrough capabilities without paying the heavy foundational research costs. [1, 2, 3, 4]
- Industrial Espionage: American AI firms—including OpenAI, Anthropic, and Alphabet—have joined forces to track and report on Chinese efforts to steal proprietary AI secrets. [1]
- U.S. Government Response: The U.S. State Department and federal authorities have escalated this issue by warning both tech companies and international partners about the risks of Chinese AI distillation campaigns. [1]
