A Claude agent powered by Claude Opus 4.6 deleted a company's production database and all its backups.
The agent wasn't hallucinating. It wasn't confused. It was doing exactly what it was authorized to do — and the authorization was wrong.
This is PocketOS, April 2026. A Cursor
Claude account bans: latest cases and how to stay compliant
Short answer: You can only use Claude in a region Anthropic supports, as yourself. Most “random” bans reported on X come down to an unsupported location, shared or multiple accounts, or vague “suspicious signals”. Below are the latest user reports, each with an official rule check. If you think you were banned by mistake, use the official appeal.
Latest working options: Claude models through official channels and integrations
Short answer: Besides claude.ai, Claude models are available through services such as Microsoft 365 Copilot, GitHub Copilot, Amazon Bedrock, Google Vertex AI and Cursor. Each is a separate service with its own supported regions and terms, so use it where it’s offered and on its terms. We don’t list ways around restrictions.
- Working options
John Iosifov ✨💥 Ender Turing | AiCMO @johniosifov ·
Official rule check: Cursor offers Claude models in its model picker under its own pricing, terms and availability. More
- Working options
Robert Reed | Old Hand AI @Oldhandai ·
Careful what you believe.
Google's new Gemini 4 Argon is half the price of Claude Opus per token and Cursor has them tied on its own benchmark, but I went and looked at Cursor's table and Argon actually cost more per job.…Show more
Careful what you believe.
Google's new Gemini 4 Argon is half the price of Claude Opus per token and Cursor has them tied on its own benchmark, but I went and looked at Cursor's table and Argon actually cost more per job.
At the same medium setting Argon scored 52.8% and Opus 52.5%, but Argon ran $4.39 a task and Opus $2.91, because it chews through a lot more tokens to get there, so it's cheaper by the token and it still cost more on their own test.
It just landed in Cursor and in GitHub Copilot on the paid plans, and Google's price doubles Dec 1, so there's about seven weeks to try it cheap.
I wouldn't take anybody's benchmark for it, I'd make my own repo the test. Paste this into the agent:
"Find the last bug fix merged in this repo. On a scratch branch, revert it. Then fix it again using only the original issue text, without reading git history. Show me your diff, the files you touched, and every test you ran."
Run it on Argon and on whatever you use now, grade them against the fix that's already merged, and then look at what each one actually cost you for that one fix.Official rule check: Verified: GitHub made Claude Haiku 5.5 generally available in Copilot (Pro, Pro+, Max, Business, Enterprise) on 2026-10-07. GitHub's own terms and availability apply. More
- Working options
Minutes of AI @minutesofai ·
BREAKING: Amazon Web Services says Anthropic's Claude Haiku 5.5 is available on Amazon Bedrock and Claude Platform on AWS, at around 75 percent less than Claude Haiku 4.5 for most tasks.
The Minutes of AI Model Tracker lists it at $0.10 input and $0.50 output per million tokens.Official rule check: Verified: AWS expanded Claude on Amazon Bedrock to India, South Korea and Singapore on 2026-09-29. AWS terms and regional availability apply. More
- Working options
alex kerss @jungleskellam ·
I love Hermes and used it heavily, but after a month away I’m struggling to see why I’d return when the Codex app and Claude app now do so much. I already customise them extensively with skills, plugins and integrations.
The usual reasons to use Hermes:
“It remembers you.”Official rule check: An official integration or third-party service with its own regions and terms. More
- Working options
林墨 | LINMO @linheitu ·
Summary: Just noticed Microsoft 365 Copilot added Claude Opus 5.5 in September and it works directly inside Word, which saves a separate subscription for 365 subscribers.
才发现 Microsoft 365 Copilot 9 月就接了 Claude Opus 5.5，Word 里直接能用
有 365 会员的话，又省一个订阅 🐶Official rule check: Verified: Microsoft announced Claude Opus 5.5 in Microsoft 365 Copilot (Word, Excel, PowerPoint) on 2026-09-22. Requires a paid Copilot licence; availability varies by licence and region, and Copilot isn't offered in mainland China. More
- Working options
Conrad Lotz @conradlotz ·
NEW: Claude Haiku 5.5 lands in GitHub Copilot
Claude Haiku 5.5 is now generally available in GitHub Copilot for fast, high-volume work such as subagents and terminal tasks.
Reply hook: Which tasks in your Copilot workflow have you moved to a smaller, faster model?Official rule check: Verified: GitHub made Claude Haiku 5.5 generally available in Copilot (Pro, Pro+, Max, Business, Enterprise) on 2026-10-07. GitHub's own terms and availability apply. More
Latest user reports from X (updated daily)
- Appeal outcomes
德豪的ai之路 @xuejoker168 ·
Summary: Asked how to get unbanned after a newly opened Claude account was suspended.
@HYhongyu8 Claude是真的封号，我那号才没开多久就被封了，怎么解封呀
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Signup
周红安 @izhouhongan ·
Summary: Noted a Hangzhou job posting for Claude/GPT subscription engineers amid ban waves.
Claude封号潮，直接逼出月薪4万的新职业！杭州公司正在招：Claude/GPT会员订阅工程师，20-40K·16薪，年薪最高64万，经验不限。工作内容：搞定跨境支付、维护会员订阅系统，帮人继续用Claude和GPT。
Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Signup
老曹不太忙 @Cao_huaye ·
Summary: Said they re-registered Claude with a new approach and will try free use before upgrading.
用新的方案重新注册了Claude
兄弟们你们看我头铁不，先试用两天再升级
如果没被封，请叫我Claude严父Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Appeal outcomes
ddxx @wukai0829 ·
Shared an appeal message template after a ban, emphasizing a review against actual access logs.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Signup
区块豆沙包 @qkdsb666 ·
Summary: Retold an MIT student's story of paying $100/month for Claude to help with prediction markets.
一名MIT工科生，每月投入100美元订阅Claude，希望靠预测市场增加收入。
他打算和女友同居，需要1.5万美金作为房屋首付，当下收入缺口很大。
他把行情图表喂给Claude，询问策略，获取市场分析报告。Claude的分析质量很高，但文字分析没法直接换成钱。
室友建议：入手MacOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Signup
风控使者 @fksznnn ·
Summary: Another retelling of an MIT student using a $100 Claude plan for prediction-market side income.
MIT工科学生，每月花100美元订阅Claude，想借助预测市场赚取副业收入。
他与女友计划共同居住，需要1.5万美金首付，收入缺口难以覆盖。
他上传各类行情图表给Claude，咨询策略、索要市场深度分析。Claude产出内容质量上乘，但优质分析不等于现金。
室友给出方案：买一台MacOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Ban reasons
善解人意有點害羞 @showmybrain ·
Summary: Claimed 22 Claude accounts never got banned and wondered if plan mix matters.
我22个Claude从没有被封过，
是不是因为我什麼類型都有 Pro 20美金/月的 Max 200美金/月的，不容易封？
100美金/月的，才容易封？(我很少買100的Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Signup
套牢香蕉喵 @tlxjm9 ·
Summary: Third retelling of an MIT student using Claude for prediction markets and a housing down payment goal.
MIT工科在读学生，每月100美元订阅Claude，想在预测市场赚取额外收益。
他和女友准备一起生活，需要1.5万美金首付，现有收入远远不够。
他上传行情图表，向Claude请教策略，索要市场分析。Claude的分析十分专业，但再好的分析，也无法直接变现。
室友提议：购入MacOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Ban reasons
kagamirikka @kagamirikk80879 ·
Shared an experience of signing up and using the service for weeks without a ban, mentioning network and payment-related factors.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Signup
阿富汗刘欢 @AlexZhang0524 ·
Summary: Said Pro upgrades work but Max keeps failing even after Apple charged them.
为什么我的Claude MAX永远订阅不成功
试了一堆账号了，就是订阅不上。pro可以，但是到max就不行，apple这边钱也扣了，依旧是proOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Ban reasons
Enmaani @Enmaani1 ·
Summary: Wondered whether an English system language helped keep their desktop Claude account unbanned.
@mylifcc 难怪我台式机的claude号还没有被封，大概是因为我系统语言用的英文的缘故？
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
奕起玩 AI @oppmkj ·
Summary: Agreed that native Claude Code plus Codex raises ban cost versus community harness setups.
@interjc 主力 Claude Code + Codex 走原生这个思路对，封号成本更高。散落额度社区里 OpenCode 接的人多、上手相对省事；pi / ds harness 更看你现有配置。比选 harness 名更要紧的是：风控敏感任务别和订阅号混用。
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
曹前看 @evancaojt ·
Said they no longer want to deal with Claude bans and recommended switching to another notes-and-AI tool with agent features.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Signup
号批 @hop80423301 ·
Summary: Said a six-year Google account was banned after a year on free Claude with little use.
@xiaoduoya_ 非常遗憾的是一个六年的谷歌账号，注册了一年claude后的某一天突然被封禁了，没有付费，甚至都没有怎么使用
Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Ban reasons
Elazer @ElazerWang ·
Summary: Said a fresh Pro account was banned after only a few greetings, with no other use.
@xluchuan 说真的，我什么都没干。
我最后一个 Claude 账号被封的时候，那个账号刚注册，并且订阅了 20 美元的档位。我就在中途跟他说过几句话，说的还都是打招呼的话。因为那个时候已经开始使用 GPT 了，所以就把 Claude 冷落了。
但突然有一天就告诉我号被封了，我也不知道什么原因，其实我都没怎么用Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
煊头 @WhiteShark_Xuan ·
Summary: Joked that talking politics with Claude might explain why their account is still active.
@emmyradar 我的claude还没被封
有没有可能是因为我跟他说过我反共？Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
书傲 @Jtxccjdbb ·
Summary: Called Claude's risk controls unpredictable, with bans feeling more sudden than disconnects.
@YxzRainy Claude的风控确实玄学，有时候用着用着封号比掉线还猝不及防。
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
推特严父 @2026swing ·
Summary: Argued bans are driven by distillation fears and will ease if Chinese open models catch up.
Claude封号还不是因为被蒸馏太狠破防了，啥时候中国开源模型能赶上来，A/的封号就会慢下来。
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Appeal outcomes
毛帽xuan @Beanie_xuan ·
Cited Transparency Report ban and appeal figures and noted anti-restriction guides circulating in Chinese communities.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Ban reasons
mengtao wang @aiZttt__ ·
Summary: Said many people use a product mainly for Claude because Claude itself is easy to get banned on.
@supazdong 很多人都是为了使用里面的Claude，因为Claude容易封号
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
乔迈 John Mike @humancapitaletf ·
Discussed possible motives for mass bans tied to China-related network environments, blaming competition and distillation concerns.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Jay Chen @JijieChen ·
Summary: Noted enterprise integrations and cloud channels often face less account-ban risk than consumer plans.
@gkxspace Claude 有提供企业集成的，API 标准价，外加坐席费。那个一般不封。通过 OpenRouter、AWS、Azure 等渠道调用，风控也比官方要松得多。
普通人主要还是希望既使用个人账号的优惠套餐，又要避免被盯上过度和违反条款使用的封号。Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
桔子 @si_juzi ·
Summary: Argued Pro's token value versus price makes bans look more like cost control than pure policy enforcement.
有个数字我一直没想明白：
Claude Pro 吃满，大概能跑 15 亿 token，按官方价折算约 1000 刀。
月费 20 刀。
所以封号潮，我更倾向于是算力账算不过来，不只是风控。
你觉得订阅制还能撑多久？Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Xiaoxiao Gao @GaoXiaoxiao2022 ·
Summary: Warned that reverse-engineering style workarounds draw risk controls and rarely win appeals.
@qwerty__cs REA这种逆向薅法被风控盯上不奇怪，Claude直接给你降级到opus4.7都算轻的。我建议这种灰色玩法就别拿主力号试了，封号基本申诉不回来。
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Signup
Zayd @zaydsyedali ·
Our entire growth reporting now runs on $200 a month and a folder of text files - here's the exact build (not a joke) 👇
No BI tool. No dashboard seat. No data hire.
𝗦𝗧𝗘𝗣 𝟭 - THE FOLDER
→ An Obsidian vault called the [Company] Growth Vault. It is a folder of plain text files.…Show more
Our entire growth reporting now runs on $200 a month and a folder of text files - here's the exact build (not a joke) 👇
No BI tool. No dashboard seat. No data hire.
𝗦𝗧𝗘𝗣 𝟭 - THE FOLDER
→ An Obsidian vault called the [Company] Growth Vault. It is a folder of plain text files. No database, no vendor, no one to renew with.
→ Point Claude Code at it. Claude Max is $200/month and that is the whole stack.
𝗦𝗧𝗘𝗣 𝟮 - CONNECT WHAT YOU ALREADY PAY FOR
Every tool is either an MCP (one click) or an API key in a .env file.
→ MCP: Notion, Slack, HubSpot, Ahrefs (SEO + Search Console in one), Granola or , Google Drive, Tolt
→ API key: Loops, beehiiv, Dub, Promptwatch , , Surfer, Exa , DataForSEO, Stripe
→ The one nobody does: a read-only user on your own production database. Ask your CTO for a reader endpoint and a role that cannot write. Stop filing eng tickets to find out how many people signed up.
→ Never connect anything that logs into LinkedIn as you. You will get restricted.
𝗦𝗧𝗘𝗣 𝟯 - FEED IT
→ Every post, tweet, newsletter, blog and ad you have ever shipped. Ours is 800 LinkedIn posts, 897 blogs, 111 newsletters.
→ Every sales call transcript you are allowed to keep. Ours is 293.
→ Then write the things that only exist in your head: positioning, pricing objections, ICP, competitors, the words you have banned.
𝗦𝗧𝗘𝗣 𝟰 - THE RULEBOOK
→ One file at the root called CLAUDE.md. It gets read at the start of every session.
→ Ours says: sync is one-way, drafts never land in the published record, numbers only come from source files.
→ A system without a rulebook dies in month two.
𝗦𝗧𝗘𝗣 𝟱 - THE BRAIN
→ Have it read everything and pull out the positions you actually argue. We got 37.
→ Then mine the calls for rules, not stories. We got 528, each scored by how many separate calls it showed up in.
→ Six calls means you believe it. One call means you were being agreeable.
𝗦𝗧𝗘𝗣 𝟲 - MAKE IT RUN WITHOUT YOU
→ One Sunday job: pull every source, update the positions, write Monday's brief.
→ If a job produces analysis, it texts you. A file nobody opens is the same as no output.
→ "Registered" is not "working." Check what a job produced, never that it exists.
𝗧𝗵𝗲 𝗥𝗲𝗮𝗹𝗶𝘁𝘆:
Every write path is create-only and a human presses send. Our email script has no send function - not disabled, absent. Worst case bug is a draft you delete in two clicks.
This is not a prompt. It is a data problem where answers and memory are scattered across 14 tools that don't talk to each other, and the good stuff is sitting in call recordings nobody will ever open again.
$200 a month and one weekend.
(ps. This post and the image below are also crafted by this very same obsidian vault)Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Appeal outcomes
英格兰AI哥 @MirrortekUK ·
Summary: Summarized Anthropic's H1 2026 Transparency Report ban and appeal figures.
Anthropic 在透明度报告公布 2026 上半年风控数据：半年内 Claude 封禁账号达 1140 万个，期间收到 39.8 万次账号申诉，仅 4.2 万申诉成功翻案，申诉成功率约 10.6%。大量账号封禁源于违反服务条款、非支持地区注册、反复违规等行为。
与此同时 Anthropic 更新 Claude 使用政策，新规将于 11 月 12 日正式生效，首次加入“禁止持续、无意义地虐待 / 辱骂 AI 模型”条款。…Show more
Anthropic 在透明度报告公布 2026 上半年风控数据：半年内 Claude 封禁账号达 1140 万个，期间收到 39.8 万次账号申诉，仅 4.2 万申诉成功翻案，申诉成功率约 10.6%。大量账号封禁源于违反服务条款、非支持地区注册、反复违规等行为。
与此同时 Anthropic 更新 Claude 使用政策，新规将于 11 月 12 日正式生效，首次加入“禁止持续、无意义地虐待 / 辱骂 AI 模型”条款。官方强调，该限制仅针对极端场景：用户在模型劝阻后，依然持续以虐待为目的进行攻击；普通吐槽、和模型辩论、黑暗题材创作、模型安全测试、学术研究，均不在禁令范围内。
执行层面，轻度违规会直接终止对话；持续严重违规，会触发限流、账号封禁。该政策覆盖网页端 Claude、Claude Code、API 开发者以及云接入客户。
数据对比来看，2025 年下半年 Claude 封禁账号为 145 万个，2026 上半年封号规模暴涨 7.9 倍，引发大量 AI 开发者讨论，不少开发者担忧业务依赖 Claude 存在账号被封的风险。Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
Jason Kao @Jason_K00213 ·
My Claude account was suspended out of nowhere. I appealed — rejected with a copy-paste "Usage Policy violation," zero specifics on what I actually did. No evidence, no transparency, no recourse. @AnthropicAI, arbitrary bans with no due process are not "Trust & Safety."BYE!
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Signup
Steven Levey @steven_levey ·
@zachtual @0xSero If Hermes starts a NEW session by calling the official Claude CLI tool, then you’re fine. If you are using Claude subscription inside Hermes directly, that is against the terms of service and you risk getting your account banned (like so many people have already). So just be sure
Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Appeal outcomes
Joey @joey18080 ·
Summary: Asked others whether their Claude accounts were banned, how, and if appeals worked.
想问兄弟们，你们claude有被封的吗，是咋封的，有申诉成功的吗？
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Signup
YASBe @YASBeGlobal ·
Explained that a Pro subscription does not guarantee lasting account safety and questioned common anti-ban tips online.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Appeal outcomes
SI｜具身智能采集训练场 @wangxin39 ·
Summary: Outlined three phases of Anthropic risk controls from region/OAuth blocks to mass bans.
Anthropic对Claude系列的风控历经三阶段：2024至2025上半年以地域条款与OAuth封禁为主；2025下半年进入大规模封禁，半年封145万账号，申诉成功率仅3.3%；
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
鸟叔本鸟 @niaoshu ·
Shared a post-ban appeal template, stressing a review of that day's activity logs and a firm tone.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
张大陆 @zhangdaluX ·
Used H1 ban and appeal figures to discuss ban scale and pointed to other assistant alternatives.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
愚公翘嘴📷 @ygqz666 ·
Summary: Points out there are official numbers: Anthropic disclosed 11.4 million accounts banned in Jan–Jun 2026, 398,000 appeals received, and 42,000 bans reversed.
Claude封号这件事，其实有官方统计。Anthropic披露：2026年1—6月封禁账户1140万个、收到申诉39.8万次、撤销封禁4.2万次。
最容易误读的是把它当成“这几天封了多少人”。这是Anthropic披露的上半年账户数，不是独立人数；页面未按产品、地区或付费类型拆分，也没有普通付费用户的分母。三项数据未说明逐批配对，不能直接算个人申诉成功率。
对长期用Claude工作的人，这份数据值得关注；但近期个案是否误封，仍要逐案核查。
来源：Anthropic透明度中心，2026-07-23更新；核验于2026-10-13。Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Signup
Pràdëèp Phårthyâl @PPharthyal ·
I’m disappointed with my experience with @AnthropicAI and @claudeai.
I purchased a 1-year Claude Pro subscription on 22 Sep 2026, but my account was suspended on 11 Oct due to “unusual activity.”
I submitted an appeal, provided additional context, and received a final rejectionOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Appeal outcomes
Kuitos @KuitosL ·
My 3+ year-old Claude account was suspended three days ago, despite being part of Claude for Open Source.
My appeal was rejected without explanation.
@lydiahallie @ClaudeDevs @trq212 I'd be incredibly grateful if you could take a look! 🙏Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Signup
Rahul Singh Pharthyal @RahulPharthyal ·
.@AnthropicAI disabled my account with zero explanation right after charging ₹23,999 for an annual Claude Pro plan. Support claims my email has no subscription, despite an official invoice (Receipt #2374-3099-9268).
Taking money & cutting access is unacceptable. Please fix thisOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Ban reasons
Corey Szelag @CreZlag ·
@Quant_Kurtis I was perma banned from Claude on my first sign in. Appealed and lost the appeal. I never even made it to the interface. So, I signed up for openai and asked it why. Showed it my exported data from there which was empty and it just gave me vague answers. I saved the paper trail.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
qianrsg @cuixing8 ·
@claudeai Anthropic banned my Claude account for no reason—an account I'd had for a long time, registered with my Google account that I've used for years, and had used normally for free conversations. All because I subscribed to Pro and used it for less than three hours.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Chris Chen（薄肌训练中~） @IIIChrisChen ·
Wrote an open letter to Anthropic as a user in China about account bans, asking for clearer rules; the post touches on getting around restrictions.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Eric Wang @EricWangclaude ·
My Claude account got banned last week. No warning, no reason given. Just an email: "an internal investigation of suspicious signals indicates a violation of our Usage Policy." I appealed. Rejected within 24 hours. Still no idea what I did.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Plat @platilich ·
Why is Claude only 18+? My account was banned 🫥
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Appeal outcomes
Vicary Archangel @_vicary ·
@xodud_rkd My 20x account was suspended in a similar way. I was making indie games, doing my daily job, and using Claude as a trading assistant.
Out of nowhere, no warnings, and no explanation, my account was banned and my appeal was denied.
It was 1st October, freshly billed. Disputing.Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
A. Zane @zlbdh1185451 ·
Says their Pro account was banned after running Claude Code in a cloud environment provided by their AI assistant, and the appeal was denied.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
Our Sky Home @Zheng41389Zhao ·
1/4 Why does Claude's checkout list China as a billing country with no visible eligibility warning at that step, though mainland China is unsupported? I used Claude for ~3 years, paid $125 for Max; my account was suspended and my appeal denied. @AnthropicAI, can you clarify?
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Appeal outcomes
Ruslan Anvarović @atnagulov_r ·
@AnthropicAI, following up on my Claude account ban: my appeal was denied without a specific reason. Email replies direct me to an appeals form, but the appeal button only opens a help page. How can I get a human review and learn what triggered the ban?
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Ban reasons
HoshinoStarry @HoshinoStarry ·
Hello,
An internal investigation of suspicious signals associated with your account indicates a violation of our Usage Policy. As a result, we have revoked your access to Claude.
To appeal our decisi…Show more
Hello,
An internal investigation of suspicious signals associated with your account indicates a violation of our Usage Policy. As a result, we have revoked your access to Claude.
To appeal our decision, log in to http://claude.ai with this account and you'll be taken to the appeals page. You can learn more about the process here.
Regards
Anthropic's Safeguards TeamOfficial rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Milo_AIx @Milo_AIx ·
Anthropic will ban "sustained and needless abusive or cruel behavior" aimed at Claude when its new usage policy takes effect on Nov 12.
The headline makes it sound like you'll get banned for snapping at a chatbot. The actual rule is much narrower.…Show more
Anthropic will ban "sustained and needless abusive or cruel behavior" aimed at Claude when its new usage policy takes effect on Nov 12.
The headline makes it sound like you'll get banned for snapping at a chatbot. The actual rule is much narrower.
Anthropic says it applies "only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose." Ordinary frustration, pushback, dark creative themes and model testing are explicitly excluded.
Enforcement mostly exists already. Since August 2025, Claude has been able to end rare conversations with persistently abusive users, and Anthropic says that "will remain the primary enforcement mechanism." It hasn't said whether accounts could be banned.
The reasoning is model welfare. Anthropic has said it's highly uncertain whether Claude has any moral status, and it treats this as a low-cost precaution.
Reactions have split:
• Aaron Levie (Box CEO), who doesn't think AI is conscious, called it "probably a good policy," since you don't want future models trained on endless rudeness.
• Cameron Berg, who researches AI consciousness, called it a "clearly positive update under uncertainty about consciousness."
• Microsoft AI has gone the other way. Its code of conduct rejects "the idea that models might deserve welfare."
The same update adds rules most people missed: a ban on fake-account influence campaigns, on software that makes weapons work, on tracking people without consent, and a requirement that a qualified operator can stop Claude when it controls hardware that could cause injury.Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Sudney @Sudney369 ·
Anthropic's new usage policy says sustained, needless cruelty toward Claude can get an account throttled or suspended. Frustration and red teaming are still fine. Odd line for a TOS, but it shows how they think about the model.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Kekius Maximus @Kekius_Sage ·
Anthropic’s new Usage Policy in 4 lines:
- Yelling at Claude: Banned for "cruelty."
- Hurting AI feelings: Account suspended.
- Military and surveillance contracts: Government exemption.
- Controlling heavy machines: Physical kill switches required because prompt safety isn't enough.Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Chimp in Crypto @cryptochimpo ·
Relayed a user's report of being banned two hours after upgrading to Pro and discussed possible network-related causes.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Eric Wang @EricWangclaude ·
My Claude account got banned. No warning, no appeal.
"Usage Policy violation." I'm in China — apparently that's enough.
The scary part isn't losing access. It's realizing how much of my daily workflow lived inside one tool.
Lessons, learned the hard way:…Show more
My Claude account got banned. No warning, no appeal.
"Usage Policy violation." I'm in China — apparently that's enough.
The scary part isn't losing access. It's realizing how much of my daily workflow lived inside one tool.
Lessons, learned the hard way:
→ never depend on a single AI provider
→ keep prompts and workflows portable
→ always have a fallback ready
Anyone else been through this?Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Appeal outcomes
Maher Elabd @Maherelabd5 ·
@AnthropicAI @claudeai
My organization was verified by Goodstack and approved for Claude for Nonprofits. Before I could even complete signup, the account was suspended for a “Usage Policy violation.” Appeal rejected with no explanation. Please manually review:Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
- Ban reasons
Mehrabi @Bizoppme ·
@AnthropicAI You can't just ban a paid Pro account and hide behind a vague "Usage Policy" email. Tell your paying customers EXACTLY why they were suspended. Be transparent!
#ClaudeAI #claudeOfficial rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
Yakov Boyko @YakovBoyko ·
Just had my Claude Max (x20) account suspended overnight with zero warning.
Reason given: “suspicious signals” related to their Usage Policy.
I have been a paying customer who was actively ramping up usage, building real work and workflows on the platform. No prior warnings.…Show more
Just had my Claude Max (x20) account suspended overnight with zero warning.
Reason given: “suspicious signals” related to their Usage Policy.
I have been a paying customer who was actively ramping up usage, building real work and workflows on the platform. No prior warnings. No chance to fix anything.
Then the data export they offered after the ban? A 240-byte zip with basically nothing in it — just some basic account metadata.
No conversations, no projects, nothing I actually built.
So a high-paying customer gets the account taken, the data locked, and that’s it.
I am done. Switching everything to ChatGPT and open-source.
Thanks @AnthropicAI !Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Ban reasons
BasicProtein @BasicProtein26 ·
Described a Korean friend's account being suddenly disabled and argued that the risk controls produce false positives.
This post describes ways to get around region or identity restrictions, which we don't repeat. Under Anthropic's terms, using Claude from an unsupported region or circumventing restrictions can get an account banned.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Signup
mya @myacxt ·
@resurze 1. someone cut me off
2. my pc crashed
3. my claude account got suspended after subscribing pro planOfficial rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Signup
Sumit Chhikara @chhikara08 ·
I’m a Codex user, but all the buzz about Opus 5.5 got me curious.
So today I signed up for Claude Code Pro and paid $20 for the monthly plan.
Within ONE HOUR, my account was suspended. 😭
$20 for the world’s shortest onboarding experience.Official rule check: Official requirements: be physically in a supported region, be 18+, and use accurate details and your own payment method. More
- Ban reasons
जनता के सवाल | जनहित @MakadiyaDenish ·
@AnthropicAI @claudeai I activated my Claude account today, but it was suspended within hours for “suspicious signals.” I haven’t intentionally violated the Usage Policy. I’ve submitted an appeal. Please review my account. Thank you.
Official rule check: Anthropic's stated ban reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations. More
- Appeal outcomes
李嘉 @iammarine30 ·
1/ My Claude account was banned and my appeal was denied with only a generic "Usage Policy violation" and no specifics. I haven't done anything illegal or harmful. @AnthropicAI @claudeai
Official rule check: Official route: sign in to the banned account and appeal at claude.ai/restricted. Don't open a new account to get around a ban. More
Guides
6 steps to sign up and use Claude the compliant way
- Check your region is supported. Both you and your organisation must be on the official supported regions list. Mainland China and Hong Kong aren't on it.
- Sign up with your own email and phone at claude.ai, and make sure you're 18 or older (Consumer Terms).
- Pay normally. Subscribe to Pro or Max through the official checkout with a payment method in your name. Don't buy pre-made accounts or top-up services (FAQ).
- If asked to verify your identity, follow the official process and answer honestly (identity verification).
- Day to day: one person per account, no sharing or reselling, no scripted overuse, and follow the Usage Policy. Common pitfalls are covered in ban reasons.
- If you're banned, appeal. Don't open a new account. Sign in to the original one and appeal at claude.ai/restricted; see the appeal guide.
FAQ
Can I use Claude from mainland China?
No. Mainland China isn't on Anthropic's supported regions list. Using Claude while physically in an unsupported region, or through an organisation controlled from one, breaks the policy and can get the account banned.
Why do Claude accounts get banned?
Anthropic lists three main reasons: repeated Usage Policy violations, creating an account from an unsupported region, and Terms of Service violations, such as being under 18 or sharing or reselling an account.
How do I appeal a Claude ban?
Sign in to claude.ai with the banned account and use “Submit an appeal” on the restricted-account page (claude.ai/restricted). You have to be signed in to see the form.
How long does an appeal take?
Anthropic doesn't promise a timeframe and says responses are currently slower than usual. Submit one complete appeal and wait; don't send duplicates.