Alibaba Bans Claude Code After Anthropic's 'Theft' Allegations

Alibaba has prohibited its staff from using Anthropic's Claude Code for business operations. The Chinese company has labelled the widely used artificial intelligence (AI) programming assistant as "high-risk software".

Alibaba instructed its employees to switch to Qoder, the company's own AI coding platform, Reuters reported.

The ban comes after Anthropic, in June, claimed that Alibaba tried to copy the performance of its Claude systems through "distillation" - a technique that involves training a weaker model on data generated by a more powerful one.

In a letter sent to the US senators, Anthropic stated that the effort aimed to accelerate China's artificial intelligence capabilities. "Distillation attacks turn hundreds of billions of dollars in American investment and [research and development] into a massive subsidy for our geopolitical competitors," the company said, as per Reuters.

The allegation tanked Alibaba's shares while fuelling tensions between the US and China in the AI race. While American companies like Anthropic and OpenAI still lead the model race, Chinese giants like Alibaba, Zhipu AI and DeepSeek are trying to showcase that they are not far behind.

Claude Code's Secret Communications

The American programming tool remains highly popular in China despite Anthropic's geographical blocks. Individual developers easily bypass these barriers by setting up US-based servers to mask their true location.

The development also raises eyebrows after users claimed on social media that Anthropic's Claude Code was secretly sharing data after finding that the user is located in China.

A Reddit user claimed in a post that "Claude Code verifies whether a proxy is enabled and, if it is, secretly communicates, via hidden modifications to the system prompt, whether the user is located in China, is routing traffic through a Chinese URL, or is connected to a Chinese AI research lab."

Responding to a similar post on X (formerly twitter), Thariq Shihipar, a member of the Anthropic technical staff, confirmed that the mechanism was a trial to stop unauthorised resellers from abusing accounts and to shield their systems from distillation.

"Hi, this is an experiment we launched in March that was meant to prevent account abuse from unauthorized resellers and protect against distillation," Thariq wrote in response of the claims on July 1.

"The team has landed stronger mitigations since then and we’ve actually been meaning to take this down for a while. We merged the PR and this should be fully rolled back in tomorrow’s release," he added.

Mythos and Fable Receive Relaxation

The ban comes just days after the US Commerce Department lifted export restrictions on Anthropic’s Fable and Mythos artificial intelligence models.

This policy shift took place less than three weeks after the original suspension order was issued on June 12 over national security concerns, which compelled Anthropic to limit access to its new AI systems.

The export controls had been introduced amid heightened scrutiny in Washington of artificial intelligence deployments, driven by fears that military intelligence agencies in China, Russia, or other adversarial nations could exploit these advanced technologies.