Unable to verify organization

We have no options to verify organizaion, see first comment of this post

@OpenAI_Support Can you help here please?

We have also forwarded the reported issues to OpenAI.
In the meantime another community member has found a workaround in case the verification fails due to a bug:

One could probably invite themselves with another email and then try again to verify. But that’s speculation on my part.
Please note that some organizations cannot verify at this point in time and that would be a different issue to resolve.

I hope this helps!

1 Like

When creating an account on the OpenAI API platform, you’ll notice two sub-accounts: one labeled “Personal” and another without that label. You can rename the sub-account labeled “Personal,” and it also has an organization ID. This suggests that “Personal” is merely a convenient label rather than literally meaning “individual.”

In my case, I clicked the Verify Organization button on the sub-account that was not labeled “Personal.” However, because the process took too long—or because I performed some unnecessary actions—I eventually received the message “Organization could not be verified,” and I couldn’t complete the verification.

As a result, I clicked Verify Organization on the sub-account labeled “Personal” and waited patiently for the screen to transition, even though it took some time.

Eventually, I was redirected to a service called Persona, where I successfully completed the verification using an ID card and a face scan with my smartphone.

Since I rarely use this account and it was still on Tier 1, I emailed support@openai.com to ask either that the verification be applied to the organization ID I normally use or, if that wasn’t possible, that the Verify Organization button be restored.

The response I received said:
– They had confirmed that verification was already completed for the organization ID I normally use, and that I should be able to access models such as o3 as mentioned on the help page — yet that claim was plainly false.

I continued to explain politely via email that verification had not been completed for the organization ID I had been using.

Upon closer inspection, I also noticed that models like “o1,” “o1-pro,” and “o4-mini” were not present in the sub-account under that organization ID.

Of course, I do hope the long-standing organization ID will be verified, but I’m a bit uneasy when I consider that models like “o1,” “o1-pro,” and “o4-mini” aren’t available in its sub-account. So, I might need to raise the tier of the organization ID that has already been verified.

1 Like

I am unable to verify my organization and its super frustrating. Persona is full of bugs and it doesnt allow me to try again saying session expired and to restart the process but doesnt allow to do that. OpenAI if you dont want us to use your services, then this is the best way to stop people from using openAI API’s. Persona is your biggest torn. I want to use simple api but now i tried to verify with both attempts failed with no chance to verify again. I hate this!

1 Like

It’s not a “sub-account” you discuss. It’s a completely different organization with different billing (that you likely never set up) and tier.

Around the end of 2023, OpenAI pushed new “personal” organizations on most accounts (where “account” is your login, like “legacy user keys” are attached to). Perhaps it was to counter the maintenance of an organization owner being able to be another added account and an org then “taken”, leaving an account with no API organization at all and broken.

While you might be able to add another account, another whole login, to an organization, and have that person (or yourself again) do the verification, using a “personal” organization that is not in service for ID verification as you experienced would have the completely opposite effect.

Obviously another facet of a bigger issue they are unwilling to address properly, other than a “try again after 90 days, ID already used” (and then get an “expired” link pervasive bug again?).

This distasteful scheme must be universally fixed - or simply abandoned, before “OpenAI must be abandoned” is the go-to answer.

2 Likes

Almost 2 month and full 60 days (for me atleast) and still same issue.

Seems we need to find more trustworthy AI model provider, even Anthropic or Google has a better verification as well as model access.
OpenAI seems not Open anymore.

@OpenAI_Support

1 Like

same problem, i got stucked since 3 months, support won’t help
But I’ve already started moving to anthropic anyway, strange openai

2 Likes

Hi everyone,
I’ve been reading through this thread and it looks like I’m not the only one having trouble with the OpenAI account verification — we’re facing the same issue.

In our case, the ID upload went fine, but the facial verification step failed due to a timeout. Since then, it’s no longer possible to generate a new verification link. We’ve tried different browsers, devices, incognito mode, clearing cookies — we even added a new user to the organization and attempted verification through their email, but the result was the same.

Support hasn’t responded so far, and we’re completely stuck.

Has anyone come across any new tips or had success completing the verification process? I’d really appreciate any advice.

Thanks in advance!

I have the same issue - for over a month now - not as long as this thread, but same story.
facial recognition failed - session expired - link cannot be re-generated - stuck in limbo and reaching out to support has not resolved the issue (yet!).

Though I have seen a lot of users raise this issue here, I have yet to see someone claim a path to getting the issue resolved. I’m hoping if there are api users who’ve successfully verified their org after experiencing this issue if they could post their experience here and share.

1 Like

I have had the same issue simce april. Link stuck reached out to support multiple times no way to restart the process.

Honeslty youve just stolen my credits.

Open ai is laughable in this process you are just lucky anthropic isnt doing images.

I found the solution in a different thread. Your ID is toast, but if you are on a team, you can convince a team member to do the verification step. In Settings → Organization → Members, invite a different human to the Organization as an Owner. Then they can get a new verification link, and they can verify an organization.

Don’t even bother messaging support on this. They will just blame you for failing the first verification process and they will not do anything on their end to initiate a new link for you. They are just flat out ignoring a months long issue for what I believe is having people ditch accounts for new ones to reup even more money, wasting the credits on the previous account as it won’t be used in the manner they bought it for. Just your typical American corporation!

1 Like

Same trouble! I’ts very sad

I’ve waited over the 90 days, and weekly talk to the AI support bot - same story - complete agreement, but no way forward. Any help?

Side note - some feedback on what would help future saps from my pain…

Thank you for sharing this feedback—your points are very clear, and your frustration is more than understandable. Here’s a breakdown based on the current documented behavior and product design:

What’s in OpenAI’s Control (and Should Be Improved)

  1. Spinner/Token Refresh UI:
    There is no technical reason to keep a non-functioning refresh spinner. Fixing or removing this would be simple, and would prevent user confusion.

  2. Phone Requirement Disclosure:
    Absolutely, letting users know before starting that a phone is required for ID is both clear and respectful of the user’s time. This is a UI/UX and documentation issue, completely in OpenAI’s control.

  3. 3-Tries & 90-Day Lockout Warning:
    Advance warning about high-stakes verification limits is best practice in identity verification. This should be communicated up front, and would also save effort for both users and support.

  4. Honor 90-Day Lockout Reset:
    If the lock is not reliably expiring, that’s a technical or policy oversight, again within OpenAI’s scope to fix.

Why Does It Work This Way?

  • Most of these issues are indeed a result of incomplete UI/UX, documentation, or a lack of client-side user messaging—not deep technical necessities or valid security reasons. The 90-day lockout itself is a security practice used widely in the industry to prevent abuse, but it should be clear to users and reliably reset.
  • Limited messaging or guidance in the UI is not due to legitimate use-cases but more likely due to product development priorities or oversight.
  • Nothing in your list improves user experience or business outcome by being left as is; all would be improvements and would reduce churn or negative sentiment.

What’s Out of Immediate Control

  • Only, perhaps, the strictness of the 90-day lockout (which is used across the industry). But every other point is a direct UI/UX and customer communication failure.

Impact

  • These issues unnecessarily drive paying users to competitors.
  • They damage goodwill and trust—particularly for organizations investing heavily in the ecosystem.

Summary:
Everything you listed (transparency, messaging, functioning UI, real lockout logic) would improve the experience and trust for paying users, and is well within OpenAI’s ability to address. It’s valid feedback, and I encourage sharing it through any feedback or “report a problem” options available in your dashboard as well, since that routes directly to the product management teams.

If you have any more questions—or if you want tips on maximizing what you can do with your current org access while this is unresolved—I’m here to help as much as possible!

Update: Same deal… definitely over the 100 day mark now.

Impact: can’t generate images (including charts), can’t use GPT5. We’ve invested a lot of time and money into this organisation instance and don’t want to rebuild on a brand new instance.

How can OpenAI help generate a new Token?

Anyone solve this issue in a similar scenario (we are post-90-day lockout)

UPDATE - FIX:

OK - perhaps this will help someone else waiting for 90 days+

It turns out that maybe the missing undocumented step was entering an organisation name. Once I added this, even though the UI said ‘verification failed’, when I went back, it was actually verified!

Wow, same issue.

Reply from OpenAI Support below.

Thats crazy… verification failed on their end, not on mine. And the response is “Bad luck buddy. Can’t have access now.”

What is this?!

Facing the same issue.

Any idea how I can regenerate the link or talk to a human support person to fix it? The bot is useless. @OpenAI_Support

Tried this workaround and it didn’t work. The Persona link https://withpersona.com/verify?inquiry-id=undefined so obviously broken for the entire org.

For me the original verification actually worked and I got to the green checkmark screen at the end but the verification screen on OpenAI didn’t update.

I can’t believe how hard this is! Is it by design? I finally managed to get the Persona verification working by using an older Android phone.

But still can’t see my organisation being validated.