Claude Account Bans, Suspensions and Appeals

A user whose Max 5x account was banned argues Anthropic may use models to read chat history and infer user identity, rather than relying only on IP and browser fingerprints.

The poster remarks that mainland Chinese users seem to get banned almost immediately on Claude and doubts anyone dares to use it.

With a main account on Pro for six months, a user asks how many days to warm up a newly registered Claude alt account before subscribing to Pro without getting banned.

The poster says unexplained Claude bans are mostly IP-related and shares a tool that checks proxy, VPN and WebRTC leaks.

Claude's autonomous personal agent Duet is rolling out US-first, and the poster worries the ban risk will be higher given past experience.

Seeing several indie developer friends' Claude accounts banned, the author argues solo builders should avoid self-hosted backends.

After their main proxy died, the poster switched to a bad node previously used to log into Claude; the account still works and they are watching for a ban.

The author jokes that with Claude bans so aggressive, Anthropic's new always-on Duet agent might wake you up with a notice that your Claude account has been banned.

The poster jokes that their Claude account was banned halfway through writing about Li Bai, calling the timing suspiciously precise.

A user reports virtual-card declines followed by their Claude account being disabled the day after a successful charge, losing all project context with no appeal response, and asks whether third-party delivery is more stable.

Refund eligibility depends on the stated ban reason; official-site subscriptions terminated for other reasons may get prorated refunds, and appeals require logging in with the banned account and exporting chats first.

A post cites figures showing Anthropic banned 11.4 million Claude accounts in the first half of 2026 alone, a sizable share of its roughly 250 million monthly active users.