Chinese technology giant Alibaba will ban employees from using Claude Code, Anthropic’s AI coding assistant, starting July 10. According to internal communications, the company has classified the tool as a high-risk application for workplace use and is directing employees to switch to its own AI coding platform, Qoder.
The move comes amid growing tensions between Alibaba and Anthropic. In recent weeks, Anthropic accused Alibaba’s Qwen team of using Claude’s outputs for model distillation—a technique that leverages one AI model’s responses to help train another. Anthropic has argued that such practices threaten its intellectual property and violate its terms of service. Following those concerns, the company also tightened restrictions aimed at preventing users in China from accessing Claude through unofficial methods.
The issue gained wider attention after developers on Reddit discovered that certain versions of Claude Code contained code designed to detect whether a user might be operating from China. The mechanism reportedly examined signals such as time zone settings, proxy configurations, and other system information.
Anthropic later confirmed that the feature had been introduced experimentally in March to help identify unauthorized accounts, unofficial resellers, and potential model distillation attempts. The company added that the system has since been refined.
Under the new policy, Alibaba employees are expected to replace Claude Code with Qoder, the company’s internally developed AI programming assistant. The decision reflects not only Alibaba’s security concerns but also a broader trend among Chinese technology companies to reduce dependence on foreign AI services and strengthen their own AI ecosystems.
The dispute highlights how competition in artificial intelligence is no longer centered solely on building the most capable models. Protecting AI technologies, preventing unauthorized model training, and navigating geopolitical restrictions have become equally important battlegrounds.
Alibaba’s decision also underscores the growing divide between the U.S. and Chinese AI industries. As governments and companies place greater emphasis on AI sovereignty and security, similar restrictions on foreign AI tools may become increasingly common.
