A site is stealing and duplicating our GPTs - how can we protect our GPTs?

My GPT was made public and I Google searched it to see if it showed in the results. It did but I noticed it is a GPT created by another author. This author: https://9.tapgpts.com/

The GPT has my exact image and description.

The website lists many GPTs. It seemes like they are copying GPTs and listing them as their own. See if they used your GPT as well.

How can we stop people doing this? Those OpenAI have any policies? Do we have any copyrights for our GPTs and prompts?


Can you actually upload a custom image for gpts? I thought they were all dalle generated. But I could be wrong here.

So if that’s the case, wouldn’t that mean that what they’re linking is actually most likely “your” “original” gpt after all, and not a copy? Of course, I could be 100% off here.

I put your and original in quotes not to disparage the concept of ownership of gpts, but to denote the difference between an instance of a copy of a class and an instance of the original class

The page 9.tapgpts.com is apparently scraping content from gptshunter.com and that page is linking to the original GPT as it is merely a directory.

So, it’s all good from the perspective of the GPT builder.

To me it looks like this page is a GPT directory, meaning they are linking to your original GPT.

Please compare the links to help us assess the situation.

Also, currently it is possible to exactly copy a existing GPT, especially if it is build only using custom prompts and uploaded files.
We are waiting for an update from OpenAI regarding this situation. Hopefully this can be partially fixed before the store is released.

Otherwise please check the forum search as there are many other topics discussing the issue of protecting your GPTs.


Yes, you can upload your own image or use DALL-E to generate the GPT picture.


In ChatGPT the copied GPTs point to 9.tapgts, not to the original GPT builder.

It seems like GPTs will fail to deliver what they promised - like plugins failed. We will have a monumental number of GPTs that no one will use especially if it is so easy to duplicate others.

I think they will be useful on an individual basis.

I tried several links and none of them linked back to the registry in question.
Can you share an example in case I am missing something here?

1 Like

OpenAI has policies in place to protect the usage of its models and the content generated by them. According to the information provided in the extracts, you own the output you create with ChatGPT, including the right to reprint, sell, and merchandise, subject to OpenAI’s Content Policy (https://labs.openai.com/policies/content-policy) and Terms (Terms of use) ^source^ (https://help.openai.com/en/articles/6783457-what-is-chatgpt).

If you believe that your GPTs are being misused or copied, you can report your concerns to OpenAI for further review. OpenAI has automatic systems in place to help ensure GPTs adhere to usage policies, preventing harmful content and impersonation ^source^ (https://help.openai.com/en/articles/8554407-gpts-faq#h_56e44d6ee1).

However, it’s important to note that while OpenAI has these policies in place, the enforcement of copyright and intellectual property rights often involves legal processes that are outside the scope of OpenAI’s control. If you believe your rights have been infringed, you may want to consult with a legal professional.


Small addendum for anyone like me who thought this was broken for the longest time that this never worked for me in firefox, for some reason. No error or anything, it just doesn’t “take”. I have to do it via a chrome-based browser.


Exactly, simple GPT’s (using the builder) are not anything special… Your GPT will be inherited protected by using Actions and connecting it with YOUR custom data sources or applications. That is where the value comes in!

Creating a prompt is not hard and low value. Focusing on unique and custom data is much more valuable and cannot be easily recreated!

The users are discussing an issue about GPTs being copied and listed by a third-party site 9.tapgpts.com. hello44 noticed that their model was among those duplicated. Diet and vb suggest that the site could be merely a directory pointing to the original GPTs, while hello44 believes that the copied GPTs don’t lead back to original creators.

vb confirms that there is an option to upload a custom image on GPTs. lli adds that this feature wasn’t previously working for them in Firefox browser. They suggest using a Chrome-based one instead.

Finally, PaulBellow shares that OpenAI has policies to protect the usage of its models and the content they generate. If users’ GPTs are being misused or copied, they can report to OpenAI. However, the enforcement of copyright rights often involves legal processes. bloodlinealphaDev points out that simple GPTs aren’t exceptional since their value comes from unique actions and connections to custom data sources or applications.

Summarized with AI on Jan 2 (GPT-4-32k)

This can be frustrating. It has happened to a few of my GPTs. In general, I have found that focusing my efforts on
a) marketing my GPTs so it floats to the top of each directory
b) continually leveling up my expertise in Instructions, Knowledge, and Actions
c) sharing publicly a lot of the techniques and findings

has been time much better spent then chasing down people who are copying the GPTs.

I know people will have different points of view her but wanted to share my experience.


I had this person also copy my GPT (gpts.tapgpts.com), including the exact name, logo, description, and starter prompts.

The incredible thing was, back in November their website was a complete copy of another person’s website (supertools.therundown.ai) including their contact information, which I found out when I emailed them). They notified AWS/Cloudflare to get the domain/site removed, so it looks like tapgpts.com just changed their site to a GPT Directory which I would not be surprised is a complete copy of someone else’s Directory.

Good news @hello44 you can go to their GPT and click “Report”. For me, I own the Trademark to my app name and logo, and it sends you to a DMCA Takedown form to fill in. After a few days, they reached out and removed the offenders GPT. This may not work if you don’t hold the copyright/trademark but there is a process and OpenAI seems to take swift action when it can.

Ideally they realize the scale this is happening (this one account has hundreds of copycat GPTs), and ban their developer account and future accounts that do this.


This does sound bad.
But I also did already ask for a example of a GPT linked by the website in question that apparently was just copied and pasted from someone else. According to my sample of around 20 GPTs I could not find such a case.

I will definitely take a look.

Not for nothing, and I’m not arguing against the logo being your IP, but you’re definitely overstating things here, given the fact your trademark filings are pending.



If you used generative AI to create it assist in the creation of the logo, then it would be an open question as to owning the copyright. And most GPT creators are likely using some type of generative AI for their logo.

I would strongly suggest people exercise caution when invoking a DMCA takedown if there does not exist a clear right to do so, as it could expose them to civil action if the claim can be demonstrated to be fraudulent.

I did not use AI to create the logo, and not overstating. People can’t steal Trademarks that are filed and pending.

Which is fine and may give you a copyright claim if you can prove you own the copyright.

It is an overstatement to say you own the trademarks though because you don’t.

I’m not looking to get into a debate about intellectual property law with you, having filed does give you the ability to pursue legal claims after your trademark is granted for any infringing uses that occurred after filing, but not right now—unless your applications have been approved and the USPTO website simply isn’t updated yet.

It would be unwise to try to use a pending trademark application to facilitate a takedown though because it’s, for whatever reason, your application is denied the victim of what would be at that point your fraudulent claim would have a good claim against you.

Regardless, my general advice is that any disputes such as this be handled through your attorneys rather than something an independent developer tries to take on themselves.

In your specific case, I saw your original post about the copied GPT and you were absolutely correct in saying they had copied everything wholesale and verbatim.

But your case is absolutely an outlier. The vast majority of GPT builders do not have trademark applications or work-for-hire logos they can claim ownership of to aid them in a takedown request.


This is just an academic note, but I found this paper from 2006:

Right, I tried to make that clear in the post that most people probably don’t have a Trademark/filing, but they can still Report the GPT and ask for it to be taken down.

The form allows for you to enter a reason, and “they copied my name, logo, description, starter prompts” seems like a good one. If OpenAI checks the timestamps and confirms who came first between two identical GPTs, I think they should have a policy to remove copycats and ban their developer account, especially if they do it hundreds of times over and piss off hundreds of authentic developers.

While your advice is well intentioned (thank you), I think the chance my trademark is denied and some copy-cat factory tries to sue me from China for defending my own business logo being re-used by them… seems pretty unlikely and I don’t see any judge ruling against me, these are free after all what claim could they use against me… seems like a stretch my friend.

I’m all for people reporting copycat GPTs, if not using trademarks and DMCA takedowns. Don’t let these people win. If all 100 ripped off developers file Reports against the same developer, I hope it wakes up OpenAI to what’s going on at scale here and ban them (over and over until the person f’s off). Who do you want to see win here?

There are multiple posters that basically act as OpenAI representatives in tone and content, as if they speak for the company. I believe that people who post like this should be far more clear that they are simply sharing their opinion as an individual. Everything about the tone and construction of this post reads as official advice by OpenAI, it is highly confusing.

I’m not sure where you’re seeing that.

1 Like