How China’s gray market sells Claude tokens at a fraction of the price

What happened

China’s strict access restrictions on Anthropic’s Claude AI model are effectively bypassed through a well-organized gray market. Despite geoblocking and selfie-based user verification designed to block Chinese users, a network called “transfer stations” resells Claude tokens at about ten percent of their official price. This workaround enables Chinese developers to tap into Claude’s capabilities without paying the full cost or meeting Anthropic’s verification requirements.

Why it matters

This gray market undermines Anthropic’s export controls and the safeguards built into their system to restrict Chinese access. It also lowers the cost for Chinese developers and firms to build AI applications with Claude, potentially accelerating AI adoption in China at a price point inaccessible through official channels. For Anthropic, this erosion of control raises both commercial risks, such as revenue loss, and safety risks, as their compliance and safety systems are effectively neutralized. For regulatory bodies, this shows how technical barriers alone may not stop AI exports or contain their use where desired.

What to watch next

Monitor if Anthropic updates its access controls to better detect and block these transfer stations or enforces stricter account verification processes. Watch for similar gray market activity around other restricted AI tools and models targeting geoblocked regions. Regulatory agencies may respond by tightening enforcement mechanisms or pushing for more robust digital export controls. Finally, AI operators and investors should track how these workarounds reshape the competitive dynamics between U.S. and Chinese AI ecosystems, especially on pricing and user access.

AI Quick Briefs Editorial Desk