Claude blocks Russian IPs: how to avoid losing your AI project

Several hundred Russian users lost access to their Claude accounts. Despite using VPN, they couldn't restore access — the server detected where the request was actually coming from. Months of work with AI agents, data, ready-made automations — all disappeared.

What happened

Anthropic blocked access from Russian IP addresses. They did it without warning: users simply tried to log in and got denied. Even if they connected through a VPN, Claude saw the real IP through browser metadata or traffic, and the block remained in effect. The exact reason is unclear. It's possible this was an Anthropic decision due to sanctions, possibly a technical error, or possibly a reaction to heavy traffic from Russia. The company issued no official statement.

What users lost

• Ready-made AI agents that wrote code and processed data
• Custom system prompts and fine-tuned configurations
• Historical conversations and context from long-term projects
• Analytical data and experimental results
• Access to payment accounts and usage history

Recovering data through Anthropic isn't working — support either doesn't respond or gives a standard "unavailable for your region" reply.

How to protect yourself

"I no longer rely on a single platform," writes one affected user.

Here's what you should do right now:

• Back up prompts and configs — save important system prompts, agent instructions in your own documents, git repositories, or cloud storage
• Export conversation histories — download Claude chats as JSON if possible, or manually rewrite key dialogues
• Use local models for critical tasks — Llama 3.1, Mistral, or open-source alternatives you can run on your own server
• Diversify your tools — work simultaneously with multiple services (Claude, GPT, Gemini, DeepSeek) and don't make them interdependent
• Set up your own API gateway — if you pay for API access, organize your own layer between your application and the service for quick switching

What this means

Claude's block in Russia shows how fragile cloud AI tools are when you rely on them completely. For freelancers and startups, this is a signal to build a more resilient tech stack: combine cloud solutions with local models, document and archive your work, don't put all your eggs in one basket. This isn't the end of Russia's AI community, but a serious reason to rethink your project architecture.