Starting July 10, Alibaba will prohibit employees from using Anthropic products in work environments and will encourage a switch to its own Qoder platform. This was reported by Pandaily and Futu.
According to their reports, the ban will affect Claude models, including Sonnet, Opus, and Fable, as well as the Claude Code tool. The company categorized it as high-risk software and recommended Qoder as an alternative. As per Reuters, the decision was driven by security concerns. Representatives from Alibaba and Anthropic did not provide comments to the agency.
What Developers Discovered
Days prior, developers noticed mechanisms in Claude Code that checked the user's environment. Among these, Reuters identified the time zone and proxy-related information.
Other sources indicated that this mechanism appeared in version 2.1.91, released in April 2026. It read data about the local time zone and checked proxy addresses or user API endpoints for keywords related to Chinese cloud providers, AI companies, and access resale services.
When a match occurred, Claude Code altered the date format and punctuation in the system prompt instead of sending explicit telemetry. This sparked discussions among developers and raised concerns about the corporate use of the tool.
Thariq Shihipar, a member of the Claude Code team at Anthropic, noted that the feature was "an experiment launched in March." He stated it was intended to combat unauthorized account resale and protect against model distillation. He also mentioned that the mechanism would be removed in the next release.
Hi, this is an experiment we launched in March that was meant to prevent account abuse from unauthorized resellers and protect against distillation.
The team has landed stronger mitigations since then and we’ve actually been meaning to take this down for a while. We merged the…
— Thariq (@trq212) June 30, 2026
Anthropic's Policy on China
Alibaba's restrictions come amid tightening policies from Anthropic regarding China and other unsupported regions. On September 4, 2025, the company updated its sales and access rules, prohibiting the use of its services in several jurisdictions due to legal, regulatory, and security risks.
Anthropic explained that organizations from such regions might be required to share data with intelligence agencies or use access to Claude to develop their own AI systems. China and Hong Kong are not listed on the supported countries and regions page.
Previously, Anthropic accused three Chinese AI startups—DeepSeek, Moonshot, and MiniMax—of a large-scale campaign to use Claude to enhance their own models. The startup claimed that the labs generated over 16 million interactions with the chatbot through approximately 24,000 fraudulent accounts, violating terms of use and regional restrictions.
In June, Anthropic reiterated its claims against Alibaba and its AI division Qwen in a letter to U.S. Senate Banking Committee Chairman Tim Scott and senior Democrat Elizabeth Warren. The developers believe this is the largest known attack of its kind, aimed at accelerating the development of Chinese models.
According to Pandaily and Futu, Anthropic's letter to the Senate triggered Alibaba's internal ban.
The Shadow Market for Access
Nonetheless, the strict restrictions have altered the demand for Claude in China. According to WIRED, accounts are being sold on Taobao, Xianyu, and Telegram, while intermediary stations have emerged for more stable operation. These services host servers in countries supported by Anthropic, purchase API access outside of China, and resell it to local users.
Anthropic representative Michael Asiman emphasized in a comment to the publication that the company employs "a set of evolving detection systems, including identity verification," and is working on identifying and shutting down proxy networks. Journalists noted an unusually high usage of Claude in Singapore relative to the country's population.
Researcher Zilan Qian from the Oxford China Policy Lab explained to WIRED that developers from China prefer Claude Code and OpenAI Codex over local alternatives.
"Chinese models are still lagging behind American ones by six to nine months, and this gap is noticeable in tasks like coding," he said.
Adding to the backdrop are U.S. export restrictions on new Anthropic models. In June, the company had to limit access to Fable 5 and Mythos 5 for foreign nationals, including its own employees outside the country.
On June 30, Anthropic announced that U.S. authorities had lifted the directive. Starting July 1, the company restored global access to Fable 5 but temporarily limited usage of the model to 50% of normal weekly limits. Mythos 5 remains available only to select American organizations.
It is worth noting that on June 24, 360 Security Technology founder Zhou Hongyi introduced an AI tool for automatically finding vulnerabilities, Tulong Feng, which he described as China's response to Anthropic's Mythos 5.