Claude Account Bans, Suspensions and Appeals

A user asks how to pay for Claude and how to keep an account from being banned.

As Anthropic bans Claude accounts en masse, the author logs into a 2013 account on an OpenAI Dots cloud VM and opens three conversations to see whether it gets banned.

The author argues Claude's new Simplified Chinese interface is a market signal only, since language options and risk control are separate systems and the same IPs and accounts will still be banned.

A commenter says a product will end up like Claude, banning accounts en masse.

A user says Claude's ban system is extremely strict and advises keeping several accounts, backing up important project files, or switching to third-party platforms.

A user reports that Claude's account bans have escalated to the point where an AI safety academic group doing non-illegal public research had its account banned.

After an update, Claude supports Simplified Chinese again, and some users worry it is a trap to lure and ban accounts, with one saying they were banned two hours after switching.

A user says they dislike Claude's targeting of Chinese users and the fear of being banned despite paying, so they choose OpenAI instead.

A seller offers Claude Max 5x and 20x top-ups and pre-made accounts from 699 yuan, promising one replacement if the account is banned, suggesting ban waves have become routine.

A user says two Claude Max accounts were banned within days because Shadowrocket ran in global proxy mode without TUN, and shares TUN-only, UDP forwarding and STUN-disabling settings to prevent it.

A user shares that buying a VPS, having Codex install a proxy on it and using it as a VPN has kept their Claude and Codex accounts unbanned for three months.

A continuously updated open-source guide lists six causes of Claude bans: data-center IPs, virtual card payments, KYC review, shared subscriptions, black-card top-ups and bot-like behavior.