Anthropic has gone to great lengths to prevent people in China from using its powerful AI model, Claude. However, Chinese users have found increasingly sophisticated workarounds, sharing their experiences on social media as soon as Fable 5 was released.

Although other Western tools like OpenAI’s ChatGPT can be accessed via Virtual Private Networks (VPNs), Anthropic has taken more aggressive steps by banning accounts believed to belong to Chinese users. Despite these measures, a thriving underground economy for Claude access thrives, with services acting as intermediaries to provide stable and reliable access.

Chinese programmers particularly favour Claude over domestic tools due to the significant gap in capabilities. However, Anthropic still does not offer commercial access to China or its subsidiaries outside the country, amid concerns about national security threats from Chinese companies.

The underground market for Claude accounts is flourishing on platforms like Taobao and Telegram, catering to both casual and heavy users who require consistent and reliable access. The ongoing cat-and-mouse game between Anthropic’s restrictions and determined users highlights the global nature of AI innovation.