US government website used AI search tool from China that FBI said copied Anthropic

The US government’s use of Alibaba’s Qwen model on a federal website reflects a disconnect in Washington’s AI strategy. The FBI accused Alibaba of "malicious" copying of Anthropic’s technology. Yet, a US agency deployed Qwen, an open-weight model, to browse public comments on regulations. This points to a lack of coordinated policy on Chinese AI tools within the US government itself.

For Asia, this incident highlights the growing scrutiny on Chinese AI models, even open-weight ones. US Representative John Moolenaar stated no federal entity should use Chinese AI. This stance will likely increase pressure on Asian companies that integrate Chinese AI tools. It creates a higher bar for trust, even when cybersecurity risks appear low, as with Qwen’s use on public data.

The test for Alibaba and other Chinese AI developers is whether they can overcome the perception of IP theft. The FBI’s accusations, though denied by China, will persist. This incident shows that even public-facing applications face immediate backlash. Watch for stricter US government guidelines on AI model procurement by late 2026. This would further limit adoption of Chinese AI in critical sectors.