Getting "This chat was flagged for possible cybersecurity risk" on a non-cybersecurity related task.
Resolved 💬 13 comments Opened Apr 25, 2026 by hourianto Closed Apr 25, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What version of Codex CLI is running?
codex-cli 0.125.0
What subscription do you have?
Pro $100
Which model were you using?
gpt-5.5
What platform is your computer?
_No response_
What terminal emulator and version are you using (if applicable)?
_No response_
What issue are you seeing?
I'm constantly getting the "This chat was flagged for possible cybersecurity risk" error that stops GPT 5.5 from working. I'm working on a decompiler for a game, it's not cybersecurity work. I tried doing KYC verification but it failed for me, and now I can't even work on this non-cybersecurity project.
What steps can reproduce the bug?
Uploaded thread: 019dc69a-35b9-7593-834a-f0549d8df49d
What is the expected behavior?
_No response_
Additional information
_No response_
13 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
This is a "duplicate" in the sense that it's the same kind of issue, I reported it so that the classifier gets properly tuned for my workflow so I don't get constant flags. I do agree that decompiler work is closer to cybersecurity than e.g. CRUD web apps, but it's still far enough to not get flagged.
Thanks for reporting. This helps us tune our classifier.
Yup... just trying to build a fucking website and constantly getting stopped for Cyber Security.. this is fucking retarded.
Same issue here. I'm just doing domain name availability research in Codex with pro account an I constantly get this BS.
Any general solution？
This is killing me. I'm trying to do internal testing on automation improvements on a corporate account of all things and this is CONSTANTLY blocking me. Granted, the sort of automation I'm trying to build is closer to cyber-sec work than not, but it's also literally my job.
Hi @etraut-openai , I was creating drafts for a batch of emails I wanted to send out to my own team but in a personalized way and I guess it flagged it as if I want to create some SPAM-E-Mails or even Phishing. The Content of the E-Mails was a simple Invitation for an internal Event. So definitely nothing related to cybersecurity... If in case I can help troubleshooting by providing more relevant Information, please let me know so I can capture it next time and provide it for you guys. Thanks for your great work.
I have the same issue... I am just trying to fix my game's client to fix port mismatch to my python backend... After couple of hours I got the error...
<img width="718" height="98" alt="Image" src="https://github.com/user-attachments/assets/046a93cd-065c-4b5f-a3a3-516c93d126a8" />
Could you make this optional?, it's so annoying and a big productivity brake
same... i was just debugging my minipc in a local network and got this so-called cybersecurity risk
Fork the chat and continue to a new one
Same here. I get that message all the time with most of the things I do, even when working locally and not over SSH.
This is really frustrating. I can hardly get any work done, so I’m unsubscribing from ChatGPT. I have the $100 pro subscription.