Anthropic is reportedly tightening its restrictions after identifying multiple methods used by Chinese companies to bypass its ban on access to Claude. Anthropic is moving to shut loopholes that have allowed Chinese companies to circumvent the AI startup's restrictions on accessing its Claude large language models.
The Financial Times found that companies including Ant Financial had accessed Anthropic's Claude AI tools through overseas subsidiaries, cloud providers and other workarounds. These findings point to a pattern where Chinese firms have used indirect routes to keep using Claude despite Anthropic's policy explicitly barring such access.
How Chinese firms reportedly bypassed the restrictions
According to the report, different companies used different tactics to get around the ban. Ant reportedly provided employees with corporate Claude accounts linked to its Singapore based entity, while ByteDance reimbursed engineers for personal Claude subscriptions that they accessed using virtual private networks.
Some companies also accessed Claude through foreign subsidiaries using cloud infrastructure, including Microsoft's Azure platform, the newspaper said. The report noted that these practices reportedly do not violate US or Chinese law, but they do breach Anthropic's terms of service, which prohibit Chinese companies and foreign entities under their control from using its AI models.
Anthropic's response and detection efforts
Anthropic has stepped up efforts to detect and shut down such access, including monitoring accounts for indicators such as users' computer time zones and targeting so called transfer station services that relay requests through overseas Claude accounts.
Responding to the report, Anthropic reiterated its existing position on China access. The company stated that it explicitly prohibits accessing or facilitating access to Claude in unsupported regions, including China, and described itself as the only frontier AI company that restricts sales to PRC controlled companies, including subsidiaries incorporated outside China. Anthropic said it enforces this policy through continuous, evolving detection systems, working alongside its partners to identify and ban accounts that violate its policies.
Context: A broader pattern of scrutiny
This is not the first time Anthropic has flagged unauthorised use of its models by Chinese entities. The company had earlier accused operators linked to Alibaba's Qwen AI lab of running a large scale distillation campaign in mid June, involving thousands of fake accounts generating tens of millions of interactions with Claude, according to reports citing a letter Anthropic sent to US senators. Anthropic has also been reported to require identity verification, including government issued IDs, for accounts flagged as suspicious.
