Alibaba Bans Claude AI Amid Distillation Dispute
A recent report has revealed that Chinese tech giant Alibaba Group has initiated a company-wide ban on all artificial intelligence products developed by the US-based startup Anthropic. The internal mandate requires employees to completely uninstall Claude-related applications, marking a significant escalation in the ongoing AI technology friction between the two nations.
According to tech media Zhidongxi’s report published on July 3, Alibaba has instructed its entire workforce to remove Anthropic products, including the Sonnet, Opus, and Fable model series, as well as the Claude Code agent. The comprehensive restriction is scheduled to take formal effect on July 10.
Prior to this ban, Alibaba had actively encouraged its developers to utilize external AI tools by offering substantial reimbursement policies since early 2026. Employees frequently used platforms like Claude, OpenAI, and Gemini, with some programmers expending hundreds of US dollars weekly. The abrupt policy reversal appears directly linked to a June 24 disclosure, wherein Anthropic informed the US Senate Banking Committee that Alibaba had allegedly utilized approximately 25,000 deceptive accounts to conduct over 28 million interactions between April and June. Anthropic characterized this activity as an "industrial-scale model distillation attack," elevating the issue to a national security concern.
The dispute coincides with broader geopolitical pressures. On June 24, Alibaba filed a lawsuit against the US Department of Defense, seeking removal from the "Chinese Military Companies List" updated on June 8. Meanwhile, Anthropic has significantly tightened its risk control measures, initiating a massive wave of unannounced account suspensions targeting Chinese users in late June. Independent developers subsequently discovered that since April 2026, Anthropic's Claude Code incorporated hidden tracking mechanisms to identify Chinese cloud providers and time zones—a surveillance tactic the Claude team later acknowledged as an experimental measure.
This mutual decoupling highlights the intensifying technological divide in the global AI sector. As US AI firms implement stricter geofencing and surveillance protocols, Chinese technology corporations are likely to accelerate their reliance on domestic large language models, potentially fragmenting the international AI development ecosystem further.