Being routed to 5.2 and disallowed to verify my account as a Pro User

Potential duplicates detected. Please review them and close your issue if it is a duplicate.

• #11883
• #11925
• #11920
• #11842
• #11867

Powered by Codex Action

I verified my ID and it worked for a day or two, then I got silently routed back to gpt-5.2. Their classifier downgrades your whole account, not just the session, and it won't reset on its own. Been stuck like this for around 24 hours. Verification doesn't matter if the classifier just routes you back again. Really disappointing for a Pro ID Verified user.

I verified my ID and it worked for a day or two, then I got silently routed back to gpt-5.2. Their classifier downgrades your whole account, not just the session, and it won't reset on its own. Been stuck like this for around 24 hours. Verification doesn't matter if the classifier just routes you back again. Really disappointing for a Pro ID Verified user.

That's unfortunate 😕

I have no doubt they're working on this as a priority. They are a great team.

But I couldn't just pretend like I'm not really bothered by how this was handled either.

That way hopefully in the future, this doesn't happen again.

I’m also being routed to 5.2, and so far OpenAI hasn’t provided any solution that actually fixes the problem. I’m verified and I’m a Pro user. This is really bothering me because all my work is stalled.

I reviewed "Being routed to 5.2 and disallowed to verify my account as a Pro User" and read all comments before proposing a patch path.

Problem I see: I am extremely frustrated at this point. I am on day 3 of being routed to a model I have 0 interest in using, for which I am paying $200 for.
Root-cause hypothesis: an auth/session state mismatch across validation steps.
First patch plan: reproduce the flow with deterministic steps, inspect auth/session state transitions, patch validation around the wrong-code path, and add regression coverage for happy and failure flows.
If this aligns with maintainer expectations, I will post repro evidence first and then open the scoped patch.

I reviewed "Being routed to 5.2 and disallowed to verify my account as a Pro User" and read all comments before proposing a patch path.

Problem I see: I am extremely frustrated at this point. I am on day 3 of being routed to a model I have 0 interest in using, for which I am paying $200 for.
Root-cause hypothesis: an auth/session state mismatch across validation steps.
First patch plan: reproduce the flow with deterministic steps, inspect auth/session state transitions, patch validation around the wrong-code path, and add regression coverage for happy and failure flows.
If this aligns with maintainer expectations, I will post repro evidence first and then open the scoped patch.

Thanks for reporting, and apologies for the inconvenience. We're aware of the issue and are actively working on a mitigation. I'll share updates as they become available.

Thanks for reporting, and apologies for the inconvenience. We're aware of the issue and are actively working on a mitigation. I'll share updates as they become available.

Thanks Mr Traut. I really appreciate everything you do.

The whole day waiting and still nothing. PLEASE @etraut-openai

I'm commenting here since the other issues have been closed as duplicates. I also verified, got unblocked, and was blocked again while doing nothing nefarious or even security related. This was 3 days ago, also. I also contacted support but have not heard back. This is on the POC individual account my company purchased for me to demonstrate before we make a company-wide buying decision.

Any update on this?

Not yet but I spoke to someone on the team and they are working on this. Please dont tag Mr Traut there's nothing he can do. I will update you guys if I hear anything. I'm quite sure they are focused on this as a priority as it is getting a fair amount of attention and I know that they care about these things. Not making excuses for it either, but hopefully this is the last time we have to go through this.

I would advise that you check your accounts again. I am now being served 5.3 again

<img width="393" height="115" alt="Image" src="https://github.com/user-attachments/assets/45dd01fa-71b2-451f-b176-228af1807a4d" />

I am pleased that it is fixed, but I still cannot upload my ID so I really want to get that fixed, but I dont think thats a github issue, thus I'm closing this. Thank you

I also too. thats crazy

Nothing here. I'm still begin routed to gpt 5.2

same here, verified my id yesterday but still being rerouted

That button doesn't even work for me. Nothing happens.

Update from Alexander:

What we've improved re requests being routed from GPT-5.3-Codex to GPT-5.2:

• We significantly reduced the number of requests and users flagged as elevated risk by calibrating our classifiers and policies. (This is what caused rerouting.) As we calibrate policies I predict this will be well under 1% of users.
• We shipped loud, per-turn notifications in product for when requests are downgraded to the CLI v0.102.0. They are coming to other clients asap.
• We resolved cases where users who verified via Trusted Access did not regain access to GPT-5.3-Codex.
• We added docs to explain this all at https://developers.openai.com/codex/concepts/cyber-safety.

Appreciate your patience these past couple days as the team worked on all of this. More improvements to come.

https://x.com/embirico/status/2023891414623592653

If you experienced a downgrade from gpt-5.3-codex to gpt-5.2 despite completing Trusted Access verification, access to gpt-5.3-codex should now be restored. This fix took longer than expected. The team needed to work through some deployment issues over the past day. The fix is now rolling out across our clusters.

If you’ve completed Trusted Access verification and are still seeing a downgrade, please open a new bug report and provide details. Likewise, if you're having trouble with the verification process, please open a bug report.

This issue is related to new cybersecurity safeguards we’re implementing for gpt-5.3-codex and future models. We’ve published additional documentation explaining these safeguards and why they’re necessary.

We’re also rolling out client updates that provide clear messaging when a downgrade occurs. The latest CLI release (v0.102.0) includes this notification. Updated versions of the Codex IDE extension and app will be available shortly with the same messaging. The notification links directly to the new documentation and verification instructions.

We recognize that this rollout caused frustration. We’ll learn from this and aim for smoother releases going forward. Thank you for your patience while we worked through this.

If you experienced a downgrade from gpt-5.3-codex to gpt-5.2 despite completing Trusted Access verification, access to gpt-5.3-codex should now be restored. This fix took longer than expected. The team needed to work through some deployment issues over the past day. The fix is now rolling out across our clusters. If you’ve completed Trusted Access verification and are still seeing a downgrade, please open a new bug report and provide details. Likewise, if you're having trouble with the verification process, please open a bug report. This issue is related to new cybersecurity safeguards we’re implementing for gpt-5.3-codex and future models. We’ve published additional documentation explaining these safeguards and why they’re necessary. We’re also rolling out client updates that provide clear messaging when a downgrade occurs. The latest CLI release (v0.102.0) includes this notification. Updated versions of the Codex IDE extension and app will be available shortly with the same messaging. The notification links directly to the new documentation and verification instructions. We recognize that this rollout caused frustration. We’ll learn from this and aim for smoother releases going forward. Thank you for your patience while we worked through this.

Thank you so much! You were great through all this. I'm really happy that it's no longer silent. That is the right move. Sorry for being a bit of a Diva about it. I really wanted to build over the weekend! Thanks for getting this resolved. You guys are great.

Same here -- re-verified twice and the backend does not authenticate. Please fix.

### What issue are you seeing? I am frustrated at this point. I am on day 3 of being routed to a model I don't want to use. When I visit chatgpt.com/cyber, and click the button. Nothing happens. The backend is rejecting verification. This is unacceptable on many levels. 1. I wasn't using it for anything nefarious 2. I wasn't alerted to the fact, but was silently re-routed 3. I wasn't told how to fix it, I had to find out on X 4. Even that didn't work. 5. I reached out to support, and they said its because I hadn't age verified. 6. Which I ALSO had to find out about after the fact. Why this isn't documented on Cyber is unknown. 7. Support has no other way to help me. Support: "If your verification attempt didn’t go through, please try again using a valid government-issued ID and ensure the photos are clear and well-lit. At this time, verification must be completed through the in-product flow, as there are no alternate verification methods available." Well what am I supposed to do when your own verification method doesnt work? There's no alternatives. I am not trying to Karen out here. I am one of your most ardent supporters, but I'm also extremely disappointed that you rolled this out in its current state without any recourse for paying customers. I've never been one to sugarcoat anything. All of the good will you're building up evaporates when you do stuff like this. I understand that you're trying to be responsible with these powerful tools, I totally get that. But silent rerouting, and no clear documentation or path to remedy is why I'm upset. I know this is unintentional, and I'm quite sure, regrettable too. I think your team is amazing and I always shout you out on X. But I think I speak for everyone who was affected by this, that you simply cannot do this to your customers/supporters, and I would appreciate it if someone would help me on this. I have reached out on discord, via support channels, and on X, and so far: not a single person has been able to help me. `` VM890:261 POST https://chatgpt.com/backend-api/compliance/cyber_verification/persona/inquiries 403 (Forbidden) reflect @ VM890:261 fetch @ VM890:481 apply @ VM890:281 reflect @ VM890:760 fetch @ VM890:980 apply @ VM890:780 c @ 2340486e-h114sxj33pi48cph.js:1 s @ 4813494d-il9ahiwx8ss4v3p0.js:2 (anonymous) @ 4813494d-il9ahiwx8ss4v3p0.js:2 t9 @ 4813494d-il9ahiwx8ss4v3p0.js:2 s @ 4813494d-il9ahiwx8ss4v3p0.js:2 (anonymous) @ 4813494d-il9ahiwx8ss4v3p0.js:2 n9 @ 4813494d-il9ahiwx8ss4v3p0.js:2 s @ 4813494d-il9ahiwx8ss4v3p0.js:2 (anonymous) @ 4813494d-il9ahiwx8ss4v3p0.js:2 (anonymous) @ 4813494d-il9ahiwx8ss4v3p0.js:2 fetch @ 4813494d-il9ahiwx8ss4v3p0.js:2 request @ 4813494d-il9ahiwx8ss4v3p0.js:2 safePost @ 4813494d-il9ahiwx8ss4v3p0.js:2 mutationFn @ 5e7688bb-nwj71fu61wq54bli.js:2 fn @ 2340486e-h114sxj33pi48cph.js:1 _ @ 2340486e-h114sxj33pi48cph.js:1 start @ 2340486e-h114sxj33pi48cph.js:1 execute @ 2340486e-h114sxj33pi48cph.js:1 await in execute mutate @ 2340486e-h114sxj33pi48cph.js:1 mutate @ 4813494d-il9ahiwx8ss4v3p0.js:2 d @ 5e7688bb-nwj71fu61wq54bli.js:2 Mt @ 5e7688bb-nwj71fu61wq54bli.js:2 ue @ 5e7688bb-nwj71fu61wq54bli.js:2 Yu @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 Pr @ 2340486e-h114sxj33pi48cph.js:1 gE @ 2340486e-h114sxj33pi48cph.js:1 fv @ 2340486e-h114sxj33pi48cph.js:1 wd @ 2340486e-h114sxj33pi48cph.js:1 cv @ 2340486e-h114sxj33pi48cph.js:1 IE @ 2340486e-h114sxj33pi48cph.js:1 OE @ 2340486e-h114sxj33pi48cph.js:1 RE @ 2340486e-h114sxj33pi48cph.js:1 FE @ 2340486e-h114sxj33pi48cph.js:1 ec @ 2340486e-h114sxj33pi48cph.js:1 `` HE @ 2340486e-h114sxj33pi48cph.js:1 (anonymous) @ 2340486e-h114sxj33pi48cph.js:1 ### What steps can reproduce the bug? It is just my account. Everyone else seems to be able to verify ### What is the expected behavior? _No response_ ### Additional information _No response_