O3 requires org verification for Tier 5 orgs which is broken

,

I am one of the owners of a Tier 5 org and I still need to verify it for accessing the o3 via API.

Problem is that I cannot regenerate the link to verify.

Is there an ETA to get this resolved?

9 Likes

Having the same issue. Very eager to try out Codex but blocked due to this

I contacted the Persona team but they were unhelpful and said to contact OpenAI. I did that yesterday through the chat bot but haven’t heard back.

Same here. Wondering if it’s a bug that can be fixed soon.

Interesting. Maybe we should ping them here.

You can try to submit a ticket to persona. Maybe if they get enough requests they’ll fix their broken software

app dot withpersona dot com/dashboard/contact-us?category=end-user

same thing happened to me. The link doesn’t refresh and keeps providing you with the same link.

I did find a temporary workaround to get a new link. I added another email address which I own as a owner to the open ai account and then used that email log in to verify the organization. A new link was generated and I went through the process again.

However..
I keep getting error message that we were unabel to verify your identity and contact support. . I tried 10x times and I keep getting the error could not verify your identity

2 Likes

Absolute garbage. Another dumb business decision. Remove this stupid limitation OpenAi.

2 Likes

Please OpenAI this is a serious issue effecting hundreds of small businesses trying to keep up. With the rate of model improvement this year, days matter when it comes to being the first to release new use cases.

1 Like

Hello,

Thanks for flagging. I've reached out to the teams working on this new verification process and am getting caught up, so forgive me if I ask any questions that don't make sense or share any off-topic info 😅 Here's some info I've gathered so far:

Access to o3 and o4-mini models is gated either by verification status or inclusion in specific trust groups.

These groups are typically composed of Tier 4 or 5 orgs that meet additional integrity requirements (e.g. account maturity, billing signal quality, etc.).

Some key points:

  • The “unable to verify” error is expected when the org trips an integrity flag, even if tier and ID verification are otherwise valid.
  • Org verification is not guaranteed to succeed for all Tier 5 accounts.
  • Manual exceptions can be made in rare cases, but we don’t have a self-serve override or retry mechanism available yet.
  • Creating a new API key may help propagate access changes more quickly, but won’t bypass the gating logic if the org doesn’t qualify.

The verification team is actively reviewing edge cases and working to improve messaging and eligibility transparency.

In the meantime, if your org should qualify, feel free to write into support@openai.com with your Org ID.

Let me know if you have any questions or if I'm completely off base here!

Thanks again.

Just adding to this as I’ve just hit this same issue.
We’re keen to try out a new feature which it seems requried orgnisation verification.

But (*due to my phones standard camera being broken, wide and front are fine) . Plus no option to verify id via webcam on my computer. I failed to verify in time, and now when i follow the link it says " Session expired - Please restart this process or request a new link to continue."

But if I try to refresh the verify link via the OpenAI page, using the refresh button to the right, it does not update the link url.

@OpenAI_Support What are the guidelines for verification? This is important for us to know since you mentioned that Tier 5 accounts are not guaranteed to verify.

Surely Anysphere, Inc. and Windsurf should qualify but if the verification will fail for them then there’s an issue with the system itself.

1 Like

@OpenAI_Support I too am unable to verify for months now :slightly_smiling_face:

We’re investigating the “session expired” and repeated failure issues with organization verification. Most recent cases fall into two categories:

  1. The same ID was used to verify a different OpenAI org, leading the attempt to get blocked. Having another eligible owner verify, or contacting Support, resolves this.
  2. Some recently upgraded orgs were in a stale state where verification wouldn’t start or complete. Again, contact Support to resolve this.

We’re also looking into reports that the “Refresh” button doesn’t always generate a new link. If you see this, or can’t verify despite Tier 5 status, please email support@openai.com with your Org ID and the admin email that attempted verification.

If you've already reached out to Support, let me know and I can track the message down myself and reply directly!

Thank you.

2 Likes

I appreciate you looking into this, but with respect, this response exemplifies the systemic issues we’re facing.

My organization (org-mMuU648VoKSWsASDWp84N2tB) has already submitted Articles of Incorporation and multiple forms of government ID to your support team multiple times. The responses I receive are either:

  1. Complete silence for weeks

  2. Copy-pasted KB articles that don’t address the actual issue

  3. Generic responses telling me to retry the same broken verification flow

Neither of your listed categories applies to my case - my ID hasn’t verified another org, and we’re not in a “stale state.” The Persona system simply rejects valid government IDs for arbitrary reasons (in my case, normal aging over 12 years).

More importantly, this goes far beyond a technical issue. OpenAI explicitly promised organizations that opt into data sharing would receive “free daily usage up to 1 million tokens across GPT-5, o3, o4-mini, and image generation models.” I opted in based on this promise, have been sharing valuable proprietary data for months, yet am blocked from these exact models due to verification requirements that were never disclosed when I enrolled in data sharing.

This constitutes a deceptive trade practice under FTC Act Section 5. You cannot induce users to share valuable data with specific promises, then retroactively gate those benefits behind undisclosed requirements. The FTC recently secured $5 billion from Meta and $391 million from Google for similar practices. Each day OpenAI continues collecting data while denying promised access potentially constitutes a separate violation subject to penalties up to $51,744.

I’ve documented everything - the promises made, the data shared, the payments accepted, and the systematic denial of services. Multiple developers across Reddit and your own forums report identical experiences. This isn’t isolated; it’s a pattern.

Hi @tim8,

Apologies for the time and effort it's taken for you to go through the verification process. We're working on improving the transparency of the process.

Looking into your org (thank you for sharing the org_id), we can see that you've been approved as of Aug 11!

If anyone in this thread is still blocked, feel free to share your org_id and I'd be happy to look into it.

Thank you!

1 Like

Hey @OpenAI_Support,

I’m having the same issue described here. I verified my personal account with my ID, and when I tried to verify my business account it failed and I can’t retry or anything. The business account is the most important to be verified, because we need access the most recent models, but we can’t without verifying. The org and the account is quite old (even before ChatGPT). Could you help me with that, please? If you need anything else, let me know.

Org_id

Thank you in advance!

Hi @murillobrand, thank you for sharing.

I believe you'll need to try with another ID or wait for 90 days:

https://help.openai.com/en/articles/10910291-api-organization-verification?utm_source=chatgpt.com#h_31fe430971

You'll need "An ID that has not been used to verify another organization in the past 90 days"

Let me know if you have any questions!
​

Probably wise not to show your Org_id in public.

Same here if you could please have a look org-NolRLOQBY648QqVISHBZ3tTh

Same here, could you please have a look org-NolRLOQBY648QqVISHBZ3tTh