English | Chinese Original

English translation of the Chinese original. This version is generated for international readers and may be refined over time.

Date: 2026-05-27

My Pro 20x account is OpenAI's grant to open source developers.

And then last night around 9. 30, the account was sealed.

Very crazy.

My first reaction was not to be angry, but to be angry.

Because I have no idea what I triggered.

I don't touch black products, I don't touch registration machines, I don't connect, I don't register accounts, I don't resell them, and I don't take them to any strange automated workflows. This account is itself a result of the sponsorship granted to the open source project, the main purpose of which is normal development, testing and code writing.

Then I woke up and I was grounded.

What is even more absurd is that two e-mails basically do not see any useful information.

The first one says my OpenAI account cannot continue to be used because of the fact that recent activity violated Terms and Usage Policies.

The second one about my OpenAI API access was also disabled, related to personal organization.

But the problem is that it did not tell me which violation it was.

No specific categories, no sample requests, no window of time and no information for review. It's a template:

We detected that your account number had activity that was inconsistent with policy.

And then the account was gone.

It's hard to accept.

Why do I think it makes no sense?

This account is the account number of GitHub's mail box, is a new number and is open-source sponsorship for OpenAI.

I don't even remember having an API key.

So if API key abuses, my first reaction is: where did this API key come from? When was it created? From what IP? What interface? Why didn't I get any earlier security warnings?

In the case of account security, the user should at least be given a clear safe disposal path.

In the case of miscalculation, there should be an explanatory, accessible and traceable process.

But the experience is:

• No early warning;
• No specific reasons;
• No evidence of irregularities;
• Lack of clear progress in the review process;
• Just give you a button to file a complaint.

This is bad for ordinary users. Those who rely on this account for development and open-source maintenance are even more ridiculous.

The possible reason I can think of

There are probably only a few reasons I can think of.

First, regional or visiting environmental controls.

I use a relatively static IP, which occasionally switch networks, but not a high frequency switch, or a proxy pool. I did check the network in my home two days ago, doing DNS, MTU, throw-out, Wi-Fi interference, mDNS, old network configuration cleanup, etc.

However, these are normal local network searches and should not be directly equated to account abuse.

Second, the system miscalculated certain login or access actions.

If that is the case, I understand that the wind control system needs to be conservative, but I don't understand why there is no middle ground. Why not a second check, a temporary lock, a password change, a review, a direct ban?

Thirdly, there is an unusual activity under an account number or organization that I do not know about.

If that is the case, then of course I would be willing to cooperate in such matters as rotation of key, withdrawal of session, inspection of authorization applications and additional security clearance. But only if I know what happened.

The problem is, I don't even know what the problem is.

Help Center's response is delicate

I went back and asked OpenAI Help Center.

The response to the client service was presumably that they could not identify the internal route directly or help me to attach the notes to the complaint. The correct path is also an appeal link in the mail, or a backup appeal form.

At the same time, they suggested that I add in my complaint:

• User ID;
• Org ID;
• Case code in two e-mail titles;
• Login error;
• Request ID;
• The approximate time window and time zone in which the problem was discovered;
• There's no abnormal billing or usage.

If the account number is suspected of being stolen, the password is changed, the equipment is withdrawn, the MFA is opened, the API key is deleted or rotated.

This set of responses is not in itself problematic.

But because it's too standard, it's more telling: The front-line customer service did not actually see the core cause, and could not even help you to attach the context to the complaint. In other words, what really makes decisions is the black box system in the back, and all users can do is stuff the material in and wait.

This is the most uncomfortable part of the experience.

You're not communicating with a clear rule system.

You're communicating with a machine that doesn't explain why.

The place that makes me sick

It's not the accounts that bother me the most.

Any big platform could have happened.

What really bothers me is this black box treatment.

An OpenAI sponsored account for an open source developer was suddenly sealed, without warning, without details and without a clear explanation. There's only a template word in the mail. There's only one login page.account_deactivated.

It would actually be very insecure.

Because you don't know what you did wrong or how to avoid next time. And even you start to wonder, is it not as long as the network environment is not "clean" and as long as the login signal is wrong, as long as the system miscalculates once, the account numbers and interests simply disappear?

It's not a development experience.

Especially for open source projects, sponsorship rights are not a toy for "you should be grateful". It enters a real stream of work, and it affects development, testing, maintenance and dissemination. If such an interest could suddenly lapse without explanation, its dependability would be given a large question mark.

What am I gonna do now?

I have prepared to go through the appeals process and submit two emails, account IDs, organizational IDs, login errors, GitHub projects, and open source sponsorship certificates.

I'll explain:

• This is Pro 20x interest sponsored by OpenAI;
• I have no registration machine, black market, transit, resale, shared access or bypass strategy;
• I don't remember driving the API Key;
• If an API activity exists, it is hoped that OpenAI will assist in confirming whether it is unauthorized;
• in the case of a miscalculation of the wind, a request for manual review;
• If there are specific problems, please at least inform them of the type of problem and of the reorientation.

I don't need the platform to open up internal rules.

But I believe that the user should be given at least a basic explanation, a reasonable recourse and an opportunity to correct the wrongs.

It would have been ridiculous:

A normal open source person gets OpenAI's open source sponsorship and the account is blocked because of an unspecified recent activity.

I'm still saying that:

It's so weird.