Alibaba reportedly bans employees from using Claude Code
TechCrunch
β’Sat, 04 Jul 2026 16:32:08 +0000
π° What Happened
Alibaba will ban its employees from using Anthropic's Claude Code programming tool starting July 10, classifying it as high-risk software. This follows reports that Anthropic had been working to block Chinese users from accessing Claude, including an experiment where Claude Code could secretly identify Chinese users to prevent unauthorized access and model distillation. Anthropic already prohibits Chinese companies from using its models. Alibaba is directing employees to use its own in-house Qoder tool instead, as the US-China technology decoupling extends deeper into the AI coding assistant space.
π The Backstory
The US-China technology war has progressively expanded from semiconductor exports and 5G into the AI domain. Anthropic β like OpenAI β restricts access to its AI models from China and Chinese-owned entities due to US export controls and concerns about intellectual property theft and model distillation (where competitors train their own models on outputs from frontier AI systems). The discovery that Claude Code had a fingerprinting mechanism to detect Chinese users sparked controversy about privacy and covert user identification. Meanwhile, Chinese tech giants like Alibaba have been rapidly developing their own AI models (like Alibaba's Qwen family and Qoder) to reduce dependence on US technology, creating parallel AI ecosystems in the two countries.
π― Why It Matters
The Claude Code ban exemplifies the accelerating bifurcation of the global AI industry into separate US and Chinese ecosystems, with national security restrictions driving both companies and countries to build fully independent AI stacks.
Alibaba will ban its employees from using Anthropic's Claude Code programming tool starting July 10, classifying it as high-risk software. This follows reports that Anthropic had been working to block Chinese users from accessing Claude, including an experiment where Claude Code could secretly identify Chinese users to prevent unauthorized access and model distillation. Anthropic already prohibits Chinese companies from using its models. Alibaba is directing employees to use its own in-house Qoder tool instead, as the US-China technology decoupling extends deeper into the AI coding assistant space.
The US-China technology war has progressively expanded from semiconductor exports and 5G into the AI domain. Anthropic β like OpenAI β restricts access to its AI models from China and Chinese-owned entities due to US export controls and concerns about intellectual property theft and model distillation (where competitors train their own models on outputs from frontier AI systems). The discovery that Claude Code had a fingerprinting mechanism to detect Chinese users sparked controversy about privacy and covert user identification. Meanwhile, Chinese tech giants like Alibaba have been rapidly developing their own AI models (like Alibaba's Qwen family and Qoder) to reduce dependence on US technology, creating parallel AI ecosystems in the two countries.
The Claude Code ban exemplifies the accelerating bifurcation of the global AI industry into separate US and Chinese ecosystems, with national security restrictions driving both companies and countries to build fully independent AI stacks.