Chinese Users Find Ways to Secretly Use Anthropic AI

While the US-based company Anthropic is attempting to strictly limit Chinese users' access to its artificial intelligence service called Claude, a large-scale underground system for bypassing these bans has emerged in the country. This was reported by Zamin.uz.

According to recent investigations, the sale of ready-made accounts, fake identities, and specialized intermediary services for system access have become an integral part of the Chinese technology market. Anthropic, founded by former employees of the open AI field, is known for its extremely serious approach to security policies.

The company has closed official access for Chinese users and is not limiting itself to just geographic blocking or network address checks. They are actively blocking suspicious profiles using hidden networks and intermediary address-changing tools.

Recently, a requirement for identity verification using government documents via special services has even been introduced. However, Chinese users have found effective ways to bypass these obstacles.

While the simplest method is using foreign phone numbers and payment cards, more complex methods involve purchasing ready-made verified accounts on trading platforms and social network channels. Cybersecurity experts note that the demand for paid subscriptions is steadily growing on the black market.

In particular, services called transit stations have become popular. These platforms are located in countries where the AI officially operates; they purchase corporate access to the company's API service and then resell it to users within China.

For the customer, this looks like a simple chat window: the request is sent to a local site, from there it goes to the system via an intermediate server, and a response is returned. Among developers and employees of technology companies, the Claude model is more valued than local similar developments.

According to researchers, Chinese specialists emphasize that national language models are lagging behind US developments by six to nine months, especially in programming issues. Therefore, such tools have become a vital necessity for complex information technology projects.

Against the backdrop of the technological competition between the US and China, Anthropic's leadership calls the access of Chinese companies to advanced models a threat to national security. In their view, Chinese companies may use the system's responses to train and develop their own models.

Recently, the American company even openly accused China's large tech giants of such practices. However, the imposed restrictions are only fueling the further development of the underground market.