Chinese e-commerce giant Alibaba Group has ordered a complete ban on employees using Anthropic's AI programming tool Claude Code starting July 10, classifying it as high-risk software, and requiring employees to switch to the group's self-developed Qoder tool. The ban covers Alibaba's global employees. According to reports from Reuters, Dow Jones and other foreign media, the reason behind it is related to concerns over "backdoor risks."
Details of Alibaba's Ban: Claude Code Listed as High-Risk Software
According to reports, Alibaba Group has officially notified global employees that starting July 10, 2026, Claude Code is prohibited and listed as high-risk software. Alibaba requires employees to switch to the group's self-developed programming tool Qoder. The ban applies to all Alibaba Group employees worldwide.
The listing of Claude Code as high-risk is a further escalation after Alibaba previously required all employees to uninstall all Claude products. Qoder is Alibaba's self-developed AI programming tool. This designation as a replacement marks Alibaba's strategy of AI self-control and controllability moving from "blocking external tools" to "fully promoting self-developed alternatives."
Claude Code Spy Code Controversy: Exposure of User Identification Mechanism
A leak on Reddit revealed that Anthropic had embedded a version in Claude Code that could secretly identify Chinese users. Anthropic's Thariq Shihipar responded on X platform, stating that this mechanism "was an experiment we launched in March to prevent unauthorized reseller account abuse and protect the model from distillation attacks."
Distillation attack refers to the practice of using the output of other AI models to train one's own model. Anthropic has previously taken legal action against Alibaba over this. Anthropic had already banned Chinese companies and Chinese-owned overseas entities from using its models; however, the controversy over the embedded user identification mechanism has further chilled Chinese tech industry trust.
Timeline of Alibaba-Anthropic Conflict: From 28.8 Million Distillation Attack Allegations to Total Ban
According to reports, the conflict between Anthropic and Alibaba has escalated since April:
This April: Anthropic publicly accused Alibaba of launching the "largest clone attack in history," using at least 28.8 million API calls to conduct distillation attacks on the Claude model, stealing model capabilities; the relationship officially broke down.
Subsequently: Alibaba ordered all employees to uninstall all Claude series products.
Recently: Claude Code was revealed to have embedded a Chinese user identification mechanism, sparking "spy code" controversy; Anthropic defended it as an experiment to prevent abuse.
July 10, 2026: Alibaba officially lists Claude Code as high-risk software, requiring all employees to switch to its self-developed Qoder.
Frequently Asked Questions
Why did Alibaba ban Claude Code at this time, and does the ban apply to global employees?
According to reports from Reuters, Dow Jones and other foreign media, Alibaba's listing of Claude Code as high-risk software is related to concerns over "backdoor risks," against the backdrop of the Claude Code spy code controversy and the distillation attack lawsuit between Anthropic and Alibaba. The ban covers all Alibaba global employees, effective July 10, 2026; specific ban details are subject to Alibaba's official internal notice.
What is a distillation attack, and how large is Anthropic's allegation against Alibaba?
A distillation attack refers to using the output of other AI models to train one's own model, essentially "copying" model capabilities without authorization. In April this year, Anthropic publicly accused Alibaba of using at least 28.8 million API calls to conduct distillation attacks on Claude, characterizing it as the "largest clone attack in history"; specific legal proceedings are subject to official statements from both parties.
What is Qoder, and why did Alibaba choose to develop its own tool instead of other alternatives?
Qoder is Alibaba Group's self-developed AI programming tool, a specific case of the Chinese tech industry's "AI self-control and controllability" strategy. Facing technology blockades by US AI companies like Anthropic against China, Alibaba chose to develop a domestic replacement; specific features and application scope are subject to Alibaba's official product announcements.
