Same issue confirmed on Windows 11 — with root-cause evidence.

What version of the Codex App are you using (From “About Codex” dialog)?

26.831.1138.0

What subscription do you have?

plus

What platform is your computer?

windows 11

What issue are you seeing?

Builds tested: 26.831.1138.0 and 26.831.1445.0 (MSIX). Symptom: clicking the icon produces no window; Task Manager shows several ChatGPT.exe with no --type=renderer child; the main process pegs one CPU core (~97%); logs stop at Appshot hotkey inactive, never reaching in_app_updates_policy_latch. Network/proxy verified fine (all OpenAI domains reachable through proxy). Profile is healthy (no crash dumps, no singleton lock).

Root-cause trigger confirmed on this machine: system EFS encryption is broken. cipher /e fails with "The request is not supported" in every location (NTFS, EFS service running, EFS certificates present, no policy block, no Event Log errors). Any copy that preserves the encrypted attribute fails with "The specified file could not be encrypted" — which matches the runtime relocation failure in the report. Package files are "Application Protected" (filecrypt driver active), and %LOCALAPPDATA%\OpenAI\Codex accumulates stale .staging-* dirs on every attempt.

What I tried that did NOT fix it: in-place re-register, full uninstall + reinstall (got the newer 26.831.1445.0, still hangs). Same as other reports.

Request: please restore the fallback that tolerates relocation failure (like 26.820.7780.0), or investigate the EFS/filecrypt interaction. Happy to provide logs.

What steps can reproduce the bug?

Builds tested: 26.831.1138.0 and 26.831.1445.0 (MSIX). Symptom: clicking the icon produces no window; Task Manager shows several ChatGPT.exe with no --type=renderer child; the main process pegs one CPU core (~97%); logs stop at Appshot hotkey inactive, never reaching in_app_updates_policy_latch. Network/proxy verified fine (all OpenAI domains reachable through proxy). Profile is healthy (no crash dumps, no singleton lock).

Root-cause trigger confirmed on this machine: system EFS encryption is broken. cipher /e fails with "The request is not supported" in every location (NTFS, EFS service running, EFS certificates present, no policy block, no Event Log errors). Any copy that preserves the encrypted attribute fails with "The specified file could not be encrypted" — which matches the runtime relocation failure in the report. Package files are "Application Protected" (filecrypt driver active), and %LOCALAPPDATA%\OpenAI\Codex accumulates stale .staging-* dirs on every attempt.

What I tried that did NOT fix it: in-place re-register, full uninstall + reinstall (got the newer 26.831.1445.0, still hangs). Same as other reports.

Request: please restore the fallback that tolerates relocation failure (like 26.820.7780.0), or investigate the EFS/filecrypt interaction. Happy to provide logs.

What is the expected behavior?

No response

Additional information

No response

No repository files, tests, or code entry points are identified in the report. Start by reproducing the launch failure with the listed MSIX builds and checking the EFS, relocation, logs, and stale staging-directory evidence. Done means the app launches despite relocation failure, or the EFS/filecrypt interaction is resolved.