Claude account disabled: the 3 causes Anthropic names, how to appeal, and the Claude Code message that is not a ban

What does it mean when Claude says your account is disabled?

Nine different messages can lock you out, and four of them aren't policy bans at all. Anthropic's appeals page sends every real ban to one form. The fix still depends on the message, so match yours first. These are the wordings in Anthropic's own docs and in public bug reports, as of October 2026:

The ban email is short. People quoting it in Anthropic's bug tracker say it reads "An internal investigation of suspicious signals associated with your account indicates a violation of our Usage Policy", then links the appeal. One May 2026 poster said it cited no clause and no conversation. That's normal, and it's why the trigger tally further down matters.

Is "This organization has been disabled" in Claude Code a ban?

Not always. In Claude Code, Anthropic's error reference gives that message a plain cause: "Claude Code is using a stale ANTHROPIC_API_KEY from a disabled Console organization. When you have a saved subscription login, the key overrides it." An old key in a shell profile or a .env file wins over a working Pro or Max login. So your claude.ai account can be fine while the terminal says it's disabled.

1. Check for the variable. Run echo $ANTHROPIC_API_KEY on macOS or Linux, or echo $env:ANTHROPIC_API_KEY in PowerShell.
2. Unset it in the current shell and delete it from your shell profile or .env file, then relaunch claude.
3. Run /status and confirm the active credential is your subscription.
4. No key set and the error stays? Then the account itself is disabled. Go to the appeal below.

Claude Code words a real suspension differently. Since version 2.1.235 it says "Your account is on hold" and links claude.ai/restricted, and before that version it showed a misleading "Login expired" instead. "Your organization has disabled Claude subscription access" is different again. Anthropic calls it "a server-side organization setting", so only your admin can change it. On 12 October 2026, 39 issues in Anthropic's claude-code repository quoted that message, and an Anthropic team member closed one in August 2026 with the same answer.

Why was my Claude account disabled?

Anthropic gives three reasons for a ban: "Repeated violations of our Usage Policy", "Account creation from an unsupported location" and "Terms of Service violations". That's the full list on its appeals page, dated 9 July 2026. Two other paths disable an account without a policy finding. An age check runs when Anthropic's systems see signs you may be under 18. And an organization you belong to can be paused "because of unusual activity" while your own account stays in good standing.

What comes right before a ban is a different question, and Anthropic doesn't answer it. So we counted. On 12 October 2026 we searched Anthropic's public claude-code repository for issues reporting a ban on a personal account or organization. We dropped feature errors and repeat posts by the same person, then sorted the 52 left by the trigger each poster named first. They run from 4 August 2025 to 7 October 2026.

Read it as a pattern in complaints, not a rate. These are Claude Code users who chose to post, and a poster's guess isn't a cause. Seven of the 23 no-trigger reports are bare Claude Code error logs with no story at all, so some of them may be the stale-key case above rather than bans. Still, three things stand out. Payment timing is the biggest named group, from 20 minutes after a gift (#68230) to a day after a $249 payment (#91411). Two of the 15 never got as far as paying: the ban hit during checkout (#40046, #53096). Anthropic names no payment rule anywhere. Nine posters tied the ban to automation, such as a plugin left running overnight or a script calling claude -p in a loop. Four named a VPN first, and three more listed one among their suspects. Anthropic's Supported Regions Policy covers use "while physically located in an unsupported region", whatever the network says. None of the 52 mentioned age, because that path runs through a Yoti email instead.

How do you appeal a disabled Claude account?

You appeal from inside the banned account. Anthropic's page says to "log in with your banned account" at claude.ai and fill out the form, and that "You must be logged in to access the appeal form." The form lives at claude.ai/restricted.

1. Log in at claude.ai with the disabled account, not a new one.
2. Export your data first. The export and delete options sit on the same screen as the form.
3. Submit the appeal once, with what you were doing when the ban hit and any payment receipt.
4. If an organization is on hold, click "Request a review" next to it instead.
5. Keep the email and its date for any refund request.

The form doesn't always load. Seven of the 52 reports say claude.ai/restricted sent them back to an ordinary claude.ai page, so no form appeared. Anthropic's page lists no second route for a ban appeal. Its only email address on that page is for disputing a warning. One July 2026 poster who emailed usersafety@anthropic.com anyway got an automated reply pointing back to the form (#75341). Retrying later, signed in, from a private window costs nothing. That's our guess, not Anthropic advice.

Was your Claude account disabled because you might be under 18?

If the email carries a Yoti link, yes: it's an age check, not a policy ban. Anthropic's age assurance article, dated 18 May 2026, says "we'll disable accounts based on indicators of minor activity." A pass reinstates the account. Yoti offers three ways to prove you're 18 or older:

• Facial age estimation: a selfie, with no ID document.
• ID verification: a photo of a passport, driving licence or national ID card.
• Yoti's Digital ID app: share a verified "over 18" attribute you already hold.

Anthropic says it receives only a pass or fail, and Yoti deletes the selfie and document images once your age is checked. The phone apps add a second route. In certain US states, new laws make the App Store and Play Store share a user's age with app makers, per Anthropic's minimum age article. Anthropic can't see or change that data, so you fix a wrong age with Apple or Google. On a family plan, a parent may have set it. If you really are under 18, no appeal helps: Anthropic's Consumer Terms require you to be 18 "or the minimum age required to consent" where you live, "whichever is higher."

How long does a Claude appeal take?

Anthropic doesn't give a time. Its appeals page opens with a warning instead: "Our response times are currently longer than normal due to our recent launch and an increase in email volume." The ban reports show the range people actually waited, each on the date the poster filed it:

Three of the 52 posters said they won their appeal. All three then hit a second wall: the login, or the organization behind it, still read as banned. If your reinstatement email arrives and the ban screen doesn't clear, reply to that email with the date it came. Don't open a fresh appeal.

Do you get a refund when Anthropic disables your Claude account?

It depends on the reason. Anthropic's Consumer Terms, effective 8 October 2025, say that if it ends your access "due to a violation of these Terms", "you will not be entitled to any refund." If it ends a subscription for any other reason, it will "refund you, on a pro rata basis, the fees you paid for the remaining portion of your Subscription after termination." The same terms let Anthropic suspend access "at any time without notice to you" if it believes you broke them.

Several ban reports describe a refund arriving on its own (#45936, #66626, #70185). Others had a refund request refused. If you paid through the App Store or Google Play, the store runs the refund, not Anthropic. Our Claude Pro refund guide covers where each request goes, and if the screen says your card failed rather than your account, the Claude card declined guide is the right page.

Can you open a new Claude account or use Claude another way after a ban?

No. Anthropic's Usage Policy bans anyone who would "Circumvent a ban through the use of a different account, such as the creation of a new account, use of an existing account, or providing access to a person or entity that was previously banned." That clause is in the version in force since 15 September 2025 and in the update that took effect on 12 November 2026. Its last part binds the apps that resell Claude too. The appeal is the only route back.

That includes Whizi, which publishes this page. Its Claude rows send your messages to Anthropic, and so does Whizi AI, which runs on Claude Haiku as of October 2026. If you're banned, use neither. What Whizi can do while you wait is narrower. It imports a Claude export, which Anthropic lets banned users download, so your old chats stay readable. You can keep working on ChatGPT Luna on Starter. Pick it by name rather than Auto, which can send a turn to a Claude model. If you'd rather close the account for good, how to delete your Claude account covers it. And if the ban came from signing up abroad, Claude not available in your country explains the 185 places where Claude is sold and the checks that decide access.