How to prove you're not illicitely impersonating someone with a GPT? (I have their permission)

If someone from OpenAI is reading this, please help!

With the GPT Store launching tomorrow, there are sets of guidelines now and one is that you can’t impersonate someone and build a GPT of their likeness without their permission…

BUT if we have their written permission and full support, how do we submit this and notify you before our GPT gets banned/removed?

We spend a lot of time building these (including cleaning and collecting data sets for the Knowledge base), and will do some marketing so don’t want to lose our link and user count (having to remake it if you remove it before checking with us, for example).



1 Like


You have a creator website link as part of your builder profile?

Besides the policies of your data retention, describe your author association. Link to author-controlled media where they acknowledge their connection.

Easily-discovered information would make any judgements fall in your favor more easily.

Where there’s a lot of “need to read between the lines”, you can do the obvious.

I suspect there won’t be much looking closely at GPTs by staff except for those with top usage placement, being featured manually, or that have been reported by users, just because of quantity.

The GPT can also be instructed to spurt out a disclaimer or information at the conclusion of its first session message only (although I’ve found that it likes to insert them in following replies anyway).

1 Like

Yes that’s what I’m worried, when he (Marc Randolph) posts about it on Social it will go viral and shoot up the app store, then they’ll notice it, check the By and see it’s not his website and delete it before checking with him or me.

I did add his website ( as a Verified Domain on my profile, but the problem is your OpenAI Dev Account can only use one at a time (and I have other GPTs for my startup under my brand WriteOnSaga with

I tried asking OpenAI to allow multiple Verified Domains across different GPTs, but they said it’s not supported and that I would need to create a second developer profile and recreate the GPT under that account (which I’d prefer not to have to do).


Hey champ!

You could possibly eliminate the problem entirely, if you can get Marc to agree to some minor rephrasing. If the GPT you’re creating is “Marc Randolph’s AI, instead of just “Marc Randolph”, it wouldn’t technically be impersonation, right?:thinking:

Thanks for the tip. I just think, if Marc asked me to build it and support it, it’s not an illicit impersonation it’s literally him and his GPT. As in, impersonation is fine, if it’s the person themselves helping make the GPT.

I just wanted a way to submit a letter of authorization from him or something, but I guess I’m hoping before OpenAI takes down GPTs they contact the author first and ask for proof/permission before removing it (based on the developer account / verified domain alone).

Based on what happened previous with plugins, I’d say that OpenAI will just unlist your GPT from the store if they think there’s a potential problem, then send you an automated email. They won’t just delete your GPT or account :laughing:

In that case people would still have access to it, and you would still be able to share the link to your GPT.

1 Like

Ok, I mean that’s not terrible because at least it gives me a chance to prove that I have permission and get it re-listed.

They just don’t have like an email address or anything, and I don’t think they read this forum much, so it’s hard to talk to a representative there.

I was able to connect with a customer support rep through their chat, but I’m not sure she’ll even be able to help, I think they just don’t have infrastructure in place to tag which GPTs have permission to do what they do (or not). They should build a list of flag or something, “blue checkmark” style.


Good idea :+1:

I don’t think you’re going to have any problems, the people getting caught by the automation, are the one’s who trigger the moderation API :sweat_smile:


Another way would be to publish the GPT using their own ChatGPT account.

It’s not impersonation if they are the one publishing it.

It’s like “Chat with an AI that’s aligned with me (and replies like me).”