Alibaba has ordered employees to stop using Anthropic’s AI-powered programming assistant, Claude Code, marking the latest escalation in the technological rivalry between the United States and China as artificial intelligence companies tighten access to advanced models amid growing national security concerns.
The Chinese technology giant will prohibit the use of Claude Code across its workforce from July 10 after classifying the software as a high-risk application, according to multiple media reports. Employees have instead been instructed to migrate to Qoder, Alibaba’s internally developed AI coding assistant, reinforcing the company’s broader strategy of reducing reliance on foreign AI technologies.
The move comes as Anthropic intensifies efforts to prevent unauthorized access to its frontier AI models from China, where U.S. export restrictions and company policies have increasingly limited the availability of advanced American AI systems.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Anthropic already bars companies based in China, as well as overseas entities owned or controlled by Chinese firms, from accessing its Claude family of AI models. The restrictions form part of broader efforts by leading U.S. AI developers to comply with American export controls aimed at limiting China’s access to cutting-edge artificial intelligence capabilities.
The latest measures suggest the company has also been closing technical loopholes that previously enabled some Chinese users to access Claude Code through unofficial channels or third-party resellers.
The issue gained public attention after a Reddit user claimed that an earlier version of Claude Code contained functionality capable of quietly identifying users located in China. The post sparked debate among developers over how AI companies are enforcing geographic restrictions.
Anthropic executive Thariq Shihipar later acknowledged the existence of the feature, describing it as part of an internal security initiative rather than user surveillance.
“This was an experiment we launched in March that was meant to prevent account abuse from unauthorized resellers and protect against distillation,” Shihipar wrote on X.
Distillation refers to the process of training a smaller AI model using the outputs generated by a more advanced model. The technique has become a major concern for leading AI developers, who argue it enables rivals to replicate sophisticated capabilities without bearing the enormous costs of developing frontier models from scratch.
Shihipar added that Anthropic has since developed stronger safeguards against unauthorized access, making the experimental detection mechanism unnecessary.
“The team has landed stronger mitigations since then and we’ve actually been meaning to take this down for a while,” he said.
The growing fragmentation of the global AI ecosystem, which has seen geopolitical tensions increasingly shaping technology choices inside major corporations, is believed to have influenced Alibaba’s decision. Rather than depending on U.S.-developed AI software, Chinese technology companies have accelerated investment in homegrown alternatives as Washington tightens restrictions on advanced semiconductors, AI chips and frontier models.
Alibaba has been investing heavily in its own AI ecosystem, developing foundation models, enterprise AI tools and coding assistants designed to compete directly with products from OpenAI, Anthropic and Google. Directing employees toward Qoder is expected to strengthen internal adoption while reducing exposure to potential disruptions arising from U.S. policy changes.
The latest development also follows mounting concerns from Anthropic over Chinese attempts to extract capabilities from its models. Earlier this year, the company accused Alibaba of carrying out what it described as the largest known AI distillation attack against Anthropic’s systems. In a letter to the U.S. Senate Banking Committee, Anthropic alleged that operators affiliated with Alibaba conducted approximately 28.8 million interactions with its AI models using roughly 25,000 fraudulent accounts over a six-week period.
Anthropic argued that stronger cooperation between industry and government is needed to prevent illicit model distillation and maintain U.S. leadership in artificial intelligence.
Alibaba has not publicly responded to those allegations.