Data was announced on the Alibaba Reuters website Chinese technology giant Alibaba banned its employees from using Claude Code, Anthropic's artificial intelligence programming tool, at work. The decision was made after the discovery of resources that can help identify users with links to China. Alibaba's ban was initially reported by Chinese media outlets and confirmed by Reuters, based on a source familiar with the matter. The ban is part of a growing dispute between the two companies. Recently, Anthropic accused Alibaba of improperly copying capabilities from the Claude AI model. The episode also reflects the increasingly intense dispute between the United States and China for leadership in the development of artificial intelligence. Claude Code is an artificial intelligence-based programming tool created by Anthropic to assist software developers. Despite restrictions imposed by the company on Chinese users and organizations, the platform has become popular among programmers in the country. According to Reuters, Alibaba employees are being instructed to use the company's own programming platform, called Qoder.
Alibaba and Anthropic did not immediately respond to Reuters' requests for comment. The Chinese company had also not publicly commented on the accusations until the publication of this report. What's behind the US blocking of Claude's AI Anthropic accuses Alibaba of copying capabilities of its AI model Anthropic said last month it was the target of a practice known as "distillation", attributed to Alibaba. The technique consists of training a less advanced artificial intelligence model based on the responses generated by a more sophisticated one. According to Anthropic, the distillation could accelerate China's efforts to reach the technological level of advanced AI systems, such as Mythos Preview, Anthropic's advanced experimental AI model. The statement is contained in a letter sent by the company to two United States senators and obtained by Reuters. The ban comes just days after developers stated that Claude Code contained features capable of collecting information about users' environment, such as time zone and internet connection settings, as well as inserting discreet markers in messages sent to Anthropic's servers. An Anthropic employee stated on Tuesday (30), in a post on the social network X, that the feature was part of an experiment launched in March to prevent the misuse of accounts by unauthorized resellers and protect the company's models against distillation practices. Also according to Reuters, the restrictions imposed by Anthropic on users in China are difficult to apply in practice, as some people can use servers located in the United States to make their connections appear to originate in that country. Still, companies have shown greater concern about legal and regulatory risks. While American AI companies try to prevent unauthorized access, resale and copying of their technologies, Chinese cloud computing and artificial intelligence companies have relied on locally developed models and open source solutions such as DeepSeek, Qwen, Moonshot and Zhipu. At the same time, Chinese AI models have been expanding their presence in the United States market, which has aroused concern among industry experts.
