The Claude language model from Anthropic has started massively blocking accounts of users from Russia. According to Baza, several hundred people have already lost access to their profiles.
Among those affected are IT specialists, developers, and entrepreneurs who used Claude for work, analytics, and project management. Many users lost knowledge bases, project architecture, and work scripts that had been built up over months, which effectively became their "external memory."
As compensation, the service refunds money for paid subscriptions, but access to the data itself cannot be restored.
Officially, Anthropic services are not available in Russia, so users connected via VPN and other methods of bypassing restrictions. AI expert Dmitry Antipov suggested that precisely such "gray" connection schemes could have caused the wave of blockages. In his opinion, the company may be testing new internal verification mechanisms and strengthening the verification of unconfirmed accounts.
AI implementation expert Roman Adamenko believes that the problem could also have been provoked by the peculiarities of VPN operation. Frequent changes in IP addresses, connection regions, and unstable connections could have been perceived by Anthropic's security systems as a violation of the platform's terms of use.