[BUG] Appeal Form Redirect Loop After Account Restriction

I have the same problem

I ran into the same problem too; how did you solve it?

I ran into the same problem too; how did you solve it?

Still haven't :(

Same issue here.

Clicked the appeal form link in the suspension email → redirects to claude.ai, no form, no way to appeal.

This broken appeal path means there is literally no recourse for users who believe their suspension is a false positive.
Requesting a working appeal channel or direct human review.

me too, why? i‘m developer

> I ran into the same problem too; how did you solve it? Still haven't :(

I might have found a useful link: https://docs.google.com/forms/d/e/1FAIpQLSdtyS8ONgH8vi6NrUpQ2lfsjSPGTZ1bW_i-E7XtGDcDZxbKeQ/viewform

me too, why? i‘m developer

This link may help: https://docs.google.com/forms/d/e/1FAIpQLSdtyS8ONgH8vi6NrUpQ2lfsjSPGTZ1bW_i-E7XtGDcDZxbKeQ/viewform

Same issue here. Clicked the appeal form link in the suspension email → redirects to claude.ai, no form, no way to appeal. This broken appeal path means there is literally no recourse for users who believe their suspension is a false positive. Requesting a working appeal channel or direct human review.

This link may help: https://docs.google.com/forms/d/e/1FAIpQLSdtyS8ONgH8vi6NrUpQ2lfsjSPGTZ1bW_i-E7XtGDcDZxbKeQ/viewform

I'm Having the same issue, my account was suspended, no way to appeals, the email I sent is automatic responses, as a developer is super useful to count with claude code, not being able to use it means not having the same power as others. It's a relly upsetting situation, happy to avoid any problem in the future, but I didn't even get a warn first.

I have the same issue

This is an official page and they have they appeal form link there, I only trust on that one:

https://support.claude.com/en/articles/14328960-identity-verification-on-claude#h_436ca483e1

This is an official page and they have they appeal form link there, I only trust on that one: https://support.claude.com/en/articles/14328960-identity-verification-on-claude#h_436ca483e1

It's that link that is not working...

I've been having the same issue, thank you @CHOCH2R for the appeal link.

I was investigating this bug and I will share more info if Anthropic responds to my email requesting permission to do so. I can confirm that the form intended to be accessed at /restricted is not a google form, so it is in my opinion unlikely that the google form mentioned in this thread is sending your data to Anthropic.

I have the same problem like video shared

I was investigating this bug and I will share more info if Anthropic responds to my email requesting permission to do so. I can confirm that the form intended to be accessed at /restricted is not a google form, so it is in my opinion unlikely that the google form mentioned in this thread is sending your data to Anthropic.

So did Anthropic reply to your email?

Same issue. I tried the google doc here without any response. https://x.com/KYLEN_OUT/status/2071949035683115357?s=20

Same here, only found this on Reddit: https://www.reddit.com/r/ClaudeAI/comments/1u23ti9/either_im_crazy_or_claudes_account_appeal_form/

The same thing happened to me this morning. My account was upgraded to Max during the night without my permission; 30 minutes later my account was suspended. And now I cannot appeal the suspension because I'm stuck with the redirect to claude.ai/new !

I did a bit more debugging on my end in case it helps.

• I captured multiple HAR files, including one while refreshing directly on https://claude.ai/restricted.
• I don't see any HTTP redirect from /restricted. The navigation appears to happen client-side.
• The first bootstrap request (/edge-api/bootstrap/.../app_start) returns 200 OK, not a 403 or another restricted-account response.
• After the app initializes, it consistently ends up on https://claude.ai/new.
• The bootstrap response includes api_disabled_reason: "trust_and_safety" and capabilities: ["chat"].
• When I try to send a message on claude.ai/new, the API returns 400 Bad Request with:

{
    "type": "error",
    "error": {
        "type": "invalid_request_error",
        "message": "This organization has been disabled.",
        "details": {
            "error_code": "organization_disabled",
            "error_visibility": "user_facing"
        }
    },
    "request_id": "req_ABC"
}

• Local Storage and Session Storage are empty, and I couldn't find any obvious cookie indicating a restricted/banned state.

From what I can tell, this looks like the frontend isn't recognizing my current account state as one that should stay on /restricted, so it falls back to /new instead. I'm not sure whether that's expected or a bug, but I figured these observations might help narrow it down.

I also captured several HAR files while reproducing the issue. They contain authentication/session information, so I haven't attached them publicly, but I'm happy to provide redacted versions or share them privately if they'd help investigate.

Edit: Small follow-up: I also checked current_user_access. It reports chat as available, along with several org/API permissions as available, even though the completion endpoint returns organization_disabled.

So the frontend seems to receive an "available" access state during initialization, but the chat API later rejects the org as disabled. This may explain why the app loads /new instead of keeping me on /restricted.

I did a bit more debugging on my end in case it helps. I captured multiple HAR files, including one while refreshing directly on https://claude.ai/restricted. I don't see any HTTP redirect from /restricted. The navigation appears to happen client-side. The first bootstrap request (/edge-api/bootstrap/.../app_start) returns 200 OK, not a 403 or another restricted-account response. After the app initializes, it consistently ends up on https://claude.ai/new. The bootstrap response includes api_disabled_reason: "trust_and_safety" and capabilities: ["chat"]. When I try to send a message on claude.ai/new, the API returns 400 Bad Request with: { "type": "error", "error": { "type": "invalid_request_error", "message": "This organization has been disabled.", "details": { "error_code": "organization_disabled", "error_visibility": "user_facing" } }, "request_id": "req_ABC" } * Local Storage and Session Storage are empty, and I couldn't find any obvious cookie indicating a restricted/banned state. From what I can tell, this looks like the frontend isn't recognizing my current account state as one that should stay on /restricted, so it falls back to /new instead. I'm not sure whether that's expected or a bug, but I figured these observations might help narrow it down. I also captured several HAR files while reproducing the issue. They contain authentication/session information, so I haven't attached them publicly, but I'm happy to provide redacted versions or share them privately if they'd help investigate.

You should be working for Anthropic.

You should be working for Anthropic.

Would love to, but I only want to be unbanned 😭

Is that Google form is from anthropic side or just collecting our data ?

Is that Google form is from anthropic side or just collecting our data ?

Honestly there's a solid chance that it didn't come from Anthropic at all. It seems weird that a tech company like Anthropic would rely on a Google form as a backup. They don't acknowledge or deny the presence of this discussion, so, it's free roam for potential scammers and data collectors.

> Is that Google form is from anthropic side or just collecting our data ? Honestly there's a solid chance that it didn't come from Anthropic at all. It seems weird that a tech company like Anthropic would rely on a Google form as a backup. They don't acknowledge or deny the presence of this discussion, so, it's free roam for potential scammers and data collectors.

I have not seen Anthropic redirect to that form in any way, and the google form has been originally shared by a random user on Reddit claiming they've got it from Anthropic support which we have no confirmation. I think it's not from Anthropic at all.

> > Is that Google form is from anthropic side or just collecting our data ? > > Honestly there's a solid chance that it didn't come from Anthropic at all. It seems weird that a tech company like Anthropic would rely on a Google form as a backup. They don't acknowledge or deny the presence of this discussion, so, it's free roam for potential scammers and data collectors. I have not seen Anthropic redirect to that form in any way, and the google form has been originally shared by a random user on Reddit claiming they've got it from Anthropic support which we have no confirmation. I think it's not from Anthropic at all.

Then how we can unbanned our accounts now

> > > Is that Google form is from anthropic side or just collecting our data ? > > > > > > Honestly there's a solid chance that it didn't come from Anthropic at all. It seems weird that a tech company like Anthropic would rely on a Google form as a backup. They don't acknowledge or deny the presence of this discussion, so, it's free roam for potential scammers and data collectors. > > > I have not seen Anthropic redirect to that form in any way, and the google form has been originally shared by a random user on Reddit claiming they've got it from Anthropic support which we have no confirmation. I think it's not from Anthropic at all. Then how we can unbanned our accounts now

Wait until they fix the /restricted link I guess

Same issue. My organization was disabled ("This organization has been
disabled."), and claude.ai/restricted immediately redirects to
claude.ai/new while logged in with the affected account, so the appeal
form is unreachable.

Emailing usersafety@anthropic.com only triggers an auto-reply linking
back to the same inaccessible form, and the thread is then auto-closed
as "no longer monitored." So there is currently no working path to
submit an appeal for org-disabled accounts.

Org ID: 39ab0767-18f7-44f4-8d93-0e3b8f000204

Same issue. My organization was disabled ("This organization has been disabled."), and claude.ai/restricted immediately redirects to claude.ai/new while logged in with the affected account, so the appeal form is unreachable. Emailing [usersafety@anthropic.com](mailto:usersafety@anthropic.com) only triggers an auto-reply linking back to the same inaccessible form, and the thread is then auto-closed as "no longer monitored." So there is currently no working path to submit an appeal for org-disabled accounts. Org ID: 39ab0767-18f7-44f4-8d93-0e3b8f000204

R.I.P. Anthropic customer service

I’ve encountered the same issue, and there appears to be no way to submit an appeal. This is quite disappointing for a company of this level.

+1 — same issue.

• Account: personal Claude Max (l**********a@gmail.com)
• Error: 400 "This organization has been disabled" on /login and claude.ai
• Console API still works (same email) — consumer org disabled, API org active
• Likely false positive: one login from Dominican Republic; normal usage from US Mac
• Appeal form redirects to claude.ai/new; usersafety@ returns only Fin auto-replies
• No human support path found yet

Workaround: Claude Code via Console API key.

+1 — same issue. Account: personal Claude Max ([leu.tejeda@gmail.com](mailto:leu.tejeda@gmail.com)) Error: 400 "This organization has been disabled" on /login and claude.ai Console API still works (same email) — consumer org disabled, API org active Likely false positive: one login from Dominican Republic; normal usage from US Mac Appeal form redirects to claude.ai/new; usersafety@ returns only Fin auto-replies No human support path found yet Workaround: Claude Code via Console API key.

I've noticed that it's mostly people with organisations that are getting this issue. I'm deleting my organisation to try again the appeal link in a few days and will update you if it works.

<img width="742" height="546" alt="Image" src="https://github.com/user-attachments/assets/b903e4ab-9162-4441-9611-00e51deaa312" />
Appeal page seems to be back for me!

<img alt="Image" width="742" height="546" src="https://private-user-images.githubusercontent.com/50145143/618416753-b903e4ab-9162-4441-9611-00e51deaa312.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.wrOPfI_f_kjliM0UspHi3ZxRcKarP7K9iH_nE3v7DiU"> Appeal page seems to be back for me!

Excellent!! i already sent the review resquest this morning too.

Yes and for anyone commenting, don't log in the normal way, go back to the email and click the link or go to claude.ai/restricted .

Going to claude.ai normally did not help (the redirect is still broken).

Yes and for anyone commenting, don't log in the normal way, go back to the email and click the link or go to claude.ai/restricted . Going to claude.ai normally did not help (the redirect is still broken).

I've directly gone for writing "claude.ai/restricted" link and it works fine

+1 — appeal denied after what I believe was an easily verifiable accident. Adding my full timeline for Trust & Safety / engineering visibility.

Account: leu.tejeda@gmail.com (personal Claude Max)

Timeline:

• ~June 27, 2026: Suspension email (Safeguards — "suspicious signals" / Usage Policy)
• Same day: API tier upgrade emails — Console API still works; only consumer org disabled
• Submitted appeal via claude.ai/restricted (two-field "Request a review" form) — status showed "Review requested"
• July 13, 2026: Appeal denied — "cannot reinstate your account at this time due to a violation of our Usage Policy" (no specific detail provided)

What I believe triggered it:
I left a Claude Code agent session running without a stop/time limit — an accidental runaway loop that spiked token usage. I stopped it once I noticed. This was not scraping, account sharing, or intentional automation — solo development on my own internal business CRM (Node.js). The usage pattern should be distinguishable from malicious abuse in your logs.

Why this feels disproportionate:

• The root cause was a beginner operational mistake (unbounded agent session), not deliberate policy evasion
• Safeguards appeal process was broken for many users (/restricted → /new redirect); I eventually got the in-product form to work
• Console API org remains active — same email, higher tier — while consumer Max is permanently denied
• I was still billed for Max during the suspension period
• Refund request via support has not been resolved

Current workaround: Claude Code via pay-as-you-go API — materially more expensive than Max for the same development workflow.

<img width="740" height="594" alt="Image" src="https://github.com/user-attachments/assets/3bd71c21-d91a-41e6-96ed-32ededc65f3c" />

Ask for Anthropic:

1. Manual re-review of usage logs for the June spike — accident vs abuse
2. Clear policy guidance: is an accidental unbounded agent loop a permanent ban with no reinstatement path?
3. Refund/credit for Max subscription during enforced downtime
4. Fix appeal + billing sync so paying subscribers aren't charged while consumer access is revoked

Happy to provide redacted screenshots (suspension email, appeal denied, "Review requested" status, billing) privately to Anthropic staff.

This isn't about evading rules — it's about proportionality and a broken appeal/refund path for paying customers.

Hey everyone,

Has anyone actually managed to get a real person from Anthropic to look at a suspension?

My first account was under my company email. I upgraded to Claude Max and about two hours later, banned. I appealed through claude.ai/restricted and it got rejected, but they didn’t explain what I had supposedly done.

I waited a few days and created a second account using my personal email. I know that probably looks bad now, but at the time I genuinely thought the first ban was some automated mistake and I needed Claude Code for work.

The second account and personal, worked fine for about a week.

I was using the official Claude Code CLI inside a Docker sandbox, working on a normal Spring Boot project. Nothing crazy. I asked it to create a class, a DB table and an API.

I went for lunch, left Claude Code running the task, came back and the second account was banned too.

My company also uses GSA for VPN access, so maybe the IP or location changes triggered something. I honestly don’t know. That’s the whole problem. Nobody tells you anything.

I emailed [usersafety@anthropic.com](mailto:usersafety@anthropic.com) and got an automated reply sending me back to the appeal page. The support chatbot did the exact same thing until I pushed it enough to finally create a support ticket.

I’m not saying I’m perfect or that nothing could possibly have triggered their system. I’m saying I want an actual person to look at the logs and tell me what happened.

Right now it feels like this:

1. Pay for Claude Max.
2. Use the official Claude Code tool.
3. Get banned and refunded.
4. Appeal.
5. Get rejected.
6. Ask for help.
7. Get sent back to the same appeal page.

That’s not a support process. That’s a loop.

Has anyone here managed to get a rejected appeal reviewed again? Did anyone actually speak to a human? And has anyone been banned while using Claude Code through Docker or a company VPN?

At this point, I don’t need another bot telling me where the appeal form is. I need someone at Anthropic to read the situation and make a proper decision.

Same class of issue: organization disabled / on hold with "Review requested" since 2026-07-28.

• Reference: TS-019fa372-7694-75fe-9c7d-a4f81f30409e
• Email: hiseiya001@gmail.com
• Organization UUID: 4e8dbfb7-aa36-4ae2-90d5-640fd065be79

I already submitted via claude.ai/restricted. The page still shows Review requested.

Status-check emails to usersafety@ only get the automated "please use the appeals form" reply. The linked trust.anthropic.com/appeals page is the enterprise Trust Center, not a consumer appeal form.

Requesting human review / confirmation that the existing request is still queued.