Hong Kong users who relied on virtual private networks to access Claude are finding themselves locked out after Anthropic quietly tightened its enforcement of regional access controls. Reports surfaced across social media and tech forums this week as users discovered their connections were being blocked, with no prior warning from the company. For a city that sits in a regulatory grey zone when it comes to US AI services, the change has landed hard.
What Changed and Who Is Affected
Claude has never been officially available in Hong Kong. Like mainland China, the territory falls outside the list of supported regions for Claude's model family. Until recently, however, many users had found a practical workaround by routing their connections through VPN servers located in supported countries. That workaround appears to have become significantly less reliable following what users describe as a sudden policy shift. Some report complete blocks; others say their sessions drop intermittently. The pattern points to more aggressive IP and traffic filtering on Anthropic's backend.
Key Facts
- Hong Kong is not an officially supported region for Claude
- Users previously accessed the service via VPN without major issues
- Anthropic has not issued a public statement explaining the tightened enforcement
- The change appears to affect both free and paid Claude users in the territory
- Reports of disruption emerged across Reddit, Telegram groups, and local tech forums
The timing is notable. Hong Kong has become an increasingly sensitive market for US technology companies navigating tensions between Washington and Beijing. Anthropic, like other American AI firms, faces legal and compliance pressure around where its technology can be deployed and who can access it. Tightening VPN enforcement is one way a company can reduce exposure without making a formal public announcement, though it does little to soften the blow for users who depend on the tool for work.
"I've been using Claude through a VPN for months without any issues. Then one morning it just stopped working entirely. No email, no warning, nothing."Hong Kong-based developer, via Reddit
A Pattern of Restricted Access in the Region
This is not the first time AI access in Hong Kong has made headlines. Financial institutions have also been drawing lines around Claude usage in the city. Earlier this year, Goldman Sachs blocked Claude access for its Hong Kong staff, and JPMorgan moved to restrict Anthropic AI access for Hong Kong employees as well. Those corporate decisions reflected compliance concerns at the firm level. Anthropic's latest move operates at the infrastructure level, cutting off a much broader group of individual users.
For many in Hong Kong, Claude had become a practical tool precisely because it offered capabilities that locally available or mainland-approved AI services do not. Losing access through VPN channels removes what was, for most, the only viable route into the platform. Whether Anthropic plans to offer any official pathway for Hong Kong users remains unclear. The company has not commented publicly on the enforcement change, and there is no indication that supported-region status for Hong Kong is under consideration in the near term. Users looking to stay current on developments can follow the latest Claude AI news as the situation evolves.
What makes the situation particularly frustrating for affected users is the absence of communication. A policy change that cuts off access overnight, with no explanation or alternative offered, leaves people with few options and little recourse. Whether this reflects a deliberate strategic decision, a response to legal counsel, or simply tighter automated enforcement of existing rules is something Anthropic has yet to clarify publicly.
