Anthropic has initiated a wave of account suspensions for users accessing its Claude AI service via VPN. The enforcement began in Hong Kong, a region not supported by the company, where users frequently employ VPNs and foreign payment data to bypass restrictions.
Affected users report receiving notifications from Anthropic’s security team citing suspicious signals as the reason for account termination. While the company has not issued a formal policy update, the actions appear to be a targeted effort to enforce regional access rules.
The crackdown follows an Anthropic report from September 2026, which accused seven Chinese AI labs of industrial distillation. The company claims that firms like Moonshot AI and DeepSeek utilized thousands of fraudulent accounts and proxy networks to train competing models.
Anthropic alleges that Moonshot AI generated over 23 million transactions between May and July 2026, masking them as originating from Singapore and Japan. The Chinese government has denied these accusations.
Russian users, who also rely on VPNs due to the lack of official support in the country, face potential risks. Although there is no widespread report of Russian account closures as of early October 2026, Anthropic previously disabled several hundred Russian profiles in May 2026.
