Codex CLI shows “join Trusted Access for Cyber” after individual TAC verification

What version of Codex CLI is running?

codex-cli 0.145.0

What subscription do you have?

ChatGPT Pro 20x

Which model were you using?

gpt-5.6-sol max

What platform is your computer?

Darwin 25.2.0 arm64 arm

What terminal emulator and version are you using (if applicable)?

tmux

Codex doctor report

What issue are you seeing?

I completed the individual Trusted Access for Cyber verification flow at https://chatgpt.com/cyber using the same ChatGPT account that I use
to sign in to Codex CLI.

However, Codex CLI repeatedly returns the following generic error during authorized defensive cybersecurity work:

Codex error: This content was flagged for possible cybersecurity risk. If this seems wrong, try rephrasing your request. To get authorized
for security work, join the Trusted Access for Cyber program: https://chatgpt.com/cyber

The problem is not that I expect all cybersecurity safeguards to be removed. The issue is that Codex continues to tell an already verified
individual user to enroll in TAC, with no indication whether this is a generic safety response, an entitlement/routing issue, or an account-
state mismatch.

Affected thread IDs: 019f94e6-a2ce-7310-a582-a69836ea7bc9

What steps can reproduce the bug?

Uploaded thread: 019f94e6-a2ce-7310-a582-a69836ea7bc9

What is the expected behavior?

A verified individual user should not receive a generic instruction to enroll in Trusted Access for Cyber again.

If a request is blocked by content-level safety controls, Codex should clearly distinguish that from TAC enrollment or account-entitlement
status. At minimum, the error should avoid implying that individual TAC verification has not been completed.

Additional information

This report concerns the accuracy and usefulness of the Codex CLI error path, not a request to bypass cybersecurity safeguards.

I understand that Trusted Access does not guarantee any particular model or remove all policy controls. The current generic enrollment

message is misleading for an already verified individual user and makes it difficult to determine whether the problem is content classification, entitlement propagation, or model routing.
My feeling is that your sensitivity to the security of my account is extremely high, which is very abnormal.