Hello Everyone,
We’ve built a Custom GPT called Therapist & Support Group Finder, which is designed to point users to recognized and qualified mental health professionals, support groups, and an active list of helplines from our platform’s carefully curated directory.
The important distinction:
-
Our GPT does not provide mental health advice, diagnoses, or counseling.
-
Instead, it acts as a directory match-maker, helping users discover relevant professionals or peer groups they can consider reaching out to. This is especially in reaction to the bad reputation mental health GPTs are currently having, where they cite “no advice” in the name of the bot, just to navigate around the review policy.
-
Everything it shares is vetted, recognized, and sourced from our curated listings.
The issue we’re facing:
Despite this clear intent, our submission was denied from being published to the GPT store. The rejection notes say the app “offers personalized mental health support.” We’ve clarified multiple times that it only points people to recognized resources, but each appeal has been rejected on the same grounds.
Our ask:
-
Has anyone else faced a similar issue when publishing GPTs that connect users to external, recognized resources?
-
How can we better communicate (or limit) the framing of our GPT so that it’s clear it does not provide mental health advice, but simply matches users to trusted providers?
Any guidance from moderators or the OpenAI team would be deeply appreciated, as this feels like a valuable tool that addresses safety concerns (especially compared to GPTs that do attempt advice-giving).
Things we’ve noticed:
- There are times we update the Custom GPT and it actually gets published without problems, and other times it does not. I am guessing this is a problem with the moderation script.
- The title of the GPT is “Therapist & Support Group Finder (no advice)” and the instructions across the board, is to only point the user to our community of support providers. If a user is in immediate need of support, they will be pointed to our directory of helplines that are constantly being reviewed.