[BUG] Pro account locked out since July 16 after incomplete Team upgrade — 28 days, zero human support response
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
Summary
A paying Pro subscriber has been completely locked out of their account for 28 days (since July 16, 2026) after initiating but not completing a Team plan upgrade. The account is confirmed active and not banned by Anthropic’s own AI support agent (Fin), yet no human has ever responded despite 30+ messages across 4 separate support threads. The user continues to be billed for a service they cannot access.
Root Cause (July 16, 2026)
Exact reproduction steps performed on the account:
- Active Pro subscription under fcmarquis@gmail.com
- Initiated upgrade to Team plan from within the Pro account
- Entered a professional email address (francois.marquis@thanalia.com) for the Team organization
- Stopped the transaction at the payment step — did not complete the upgrade
- After this, login on all non-Apple platforms redirects to onboarding as if the account no longer exists
- The iPhone app continued working on a cached session token (confirming account data still exists server-side)
This is a plan-state mapping failure: the incomplete Team upgrade left the account in a limbo state — Pro is still active, Team was never finalized, but the auth/routing layer sends the user to onboarding instead of their existing Pro workspace.
Current Status (August 12, 2026)
• Completely locked out — cannot log in on web (any browser), desktop, or mobile (cached iOS session has since expired)
• Fin AI agent has repeatedly confirmed: Pro plan is active, account is NOT banned or suspended
• User has been billed at least once (July 24 invoice received) for a service they cannot access
• No human agent has ever responded across any channel
Complete Support Timeline (28 days)
Thread 1 — “Perte de mon compte” (July 16–Aug 3, 2026)
Conversation ID: 215475119275449
Thread 2 — “Suivi du dossier” (July 21, 2026)
• User opens separate thread to ask about progress. Fin redirects to original thread.
Thread 3 — “Demande urgente de déblocage” (August 12, 2026, French)
• User writes in French requesting urgent unblock. Fin responds with standard suspension reasons — completely misdiagnoses as a suspension when it’s a technical lockout.
Thread 4 — “Account suspended” to usersafety@anthropic.com (August 12, 2026)
• User contacts User Safety team. Safeguards Team assumes account is banned, sends link to appeals form.
• Appeals form requires login — which is the exact problem being reported (catch-22).
• User attempts to clarify. Ticket auto-closed within 31 seconds: “This conversation has been closed and is no longer monitored.”
Thread 5 — “Account locked out” to support@anthropic.com (August 12, 2026)
• User provides complete timeline in English. Fin confirms Pro plan is active, not banned.
• User asks Fin to escalate. Fin admits: “As an AI agent, I’m not able to flag your account for investigation or take direct action.”
• Fin recommends: “Sign in and click Get Help” — the exact thing the user cannot do.
What Fin Has Confirmed
Across multiple threads, Fin has repeatedly stated:
• The Pro plan (fcmarquis@gmail.com's Organization) is active
• The account is NOT banned or suspended
• The issue is a technical/account-state problem, not a policy action
• The issue requires human intervention — Fin cannot resolve it
• The support team has access to the full conversation history
Yet in 28 days, no human has ever responded.
Related Issues
This appears to be a server-side identity/plan-state mapping failure similar to:
• #68861 — magic-link verification failure, 1+ month unresolved, no human response
• #60022 — onboarding redirect after auth failure (described as backend identity-mapping issue)
• #79808 — locked out of paid account, no escalation path through Fin
Per #60022 and TrustPilot reviews, this is a systemic pattern: paying users locked out with no way to reach a human.
Impact
• 28 days without access to a paid Pro account
• Billed during lockout (at least one invoice received July 24)
• Months of critical conversation history inaccessible — used for startup development (architectural decisions, code, business strategy)
• Startup team members blocked waiting for outputs that depend on Claude context
• ~30 support messages sent across 5 threads with zero human response
• All support channels are circular: every path to get help requires the login access that is broken
What the Fix Likely Is
Based on Fin’s own diagnosis: the incomplete Team upgrade with francois.marquis@thanalia.com left a dangling organization state that conflicts with the existing Pro organization under fcmarquis@gmail.com. The routing layer sees the incomplete Team setup and redirects to onboarding instead of the Pro workspace. A backend engineer needs to either:
- Remove/clean up the incomplete Team organization, OR
- Complete the Team→Pro rollback and restore the account routing
This is likely a 5-minute fix for someone with database access.
What Should Happen?
• Login access restored to the Pro account under fcmarquis@gmail.com
• Billing credit for the month(s) of inaccessibility
• The incomplete Team plan state cleaned up
Error Messages/Logs
On login attempt (all non-Apple platforms):
• Redirected to onboarding/setup flow
• Message: “Complete your subscription” (in French: “Terminer votre abonnement”)
• No way to bypass to existing Pro workspace
Steps to Reproduce
- Have an active Pro subscription under an email address
- From within the Pro account, initiate an upgrade to a Team plan
- Enter a different professional email address for the Team organization
- Stop the transaction at the payment step — do NOT complete the upgrade
- Attempt to log in again → redirected to onboarding/setup flow with message "Complete your subscription" instead of existing Pro workspace
- Account is now inaccessible on all platforms except devices with cached sessions (iOS app)
Claude Model
Sonnet (default)
Is this a regression?
I don't know
Last Working Version
_No response_
Claude Code Version
n/a
Platform
Anthropic API
Operating System
Windows
Terminal/Shell
Other
Additional Information
Platform
claude.ai (all browsers: Firefox, Chrome, Edge), Claude Desktop (Windows, Linux), iOS app (session expired), iPad app
Operating System
Linux (Ubuntu), Windows, iOS, iPadOS, ChromeOS
Additional Information
• Anthropic Support Conversation ID: 215475119275449
• Account email: fcmarquis@gmail.com
• Team upgrade email used: francois.marquis@thanalia.com
• Pro plan confirmed active by Fin AI agent on multiple occasions
• User is founder of a Quebec-based startup (Thanalia inc.) and relies on Claude conversation history for product development
• Happy to verify identity in any way required
• Requesting human engineer review — the AI support loop has been exhaustively attempted and cannot resolve this class of issue
This issue has 1 comment on GitHub. Read the full discussion on GitHub ↗