MOST NEWS

Independent Publishers Association

Support us

Anthropic began blocking Claude accounts for users in Russia and Hong Kong

Private draft

On October 1, Anthropic began blocking Claude accounts for users from Russia and Hong Kong. In its letters, the company explains this by “suspicious signals” and offers to file an appeal.

This was reported by The Bell. In its letters, Anthropic says it revoked access because of “suspicious signals” indicating a violation of the service rules, and offers to file an appeal. One of these letters was published in his Telegram channel by Roman Maresov, head of e-commerce and transactional AI at Yandex.

The author of the Telegram channel ai_volution wrote about the same letter, and the complaints from his subscribers were noticed by the Russian outlet vc.ru. According to one user, he had been paying for a subscription at $200 per month for more than a year. Another said he accessed Claude through a virtual server in Helsinki.

In Hong Kong, complaints appeared the same day on X and Reddit, writes Hong Kong’s South China Morning Post. Among those affected were people who had long connected to Claude through a VPN, choosing, for example, Singapore, and who provided foreign payment details when registering. SCMP notes that Anthropic did not announce any change in the rules and did not respond to the outlet’s request.

According to the company, Russia, mainland China, Hong Kong, and Macau are not included in the list of countries where Claude operates. The outlet believes Anthropic may have tightened enforcement of long-standing restrictions, since the number of blocks has risen sharply. Such waves have been recurring since spring: in April, Anthropic introduced passport and selfie identity verification for some users, and in May the Telegram channel Baza reported that “several hundred” Russian users lost their accounts, with their subscription fees refunded.

In July, the British business newspaper Financial Times wrote that the company was closing loopholes through which Chinese companies used its models, including overseas subsidiaries, cloud services, and VPNs. In September, Anthropic accused seven Chinese AI labs of distillation — training their own models on Claude’s responses. According to the company, Moonshot AI and DeepSeek bypassed regional bans using thousands of fake accounts and proxy networks, and Moonshot accounts appeared to be registered in Singapore and Japan. At the time, Anthropic said it prohibits access to Claude from unsupported regions and continuously updates systems that detect and block violators.

Sources

This post is available in the following languages:

Закажи IT-проект, поддержи независимое медиа

Часть дохода от каждого заказа идёт на развитие МОСТ Медиа

Заказать проект
Link