I would like to suggest a relatively simple feature that could improve the feedback process for both ChatGPT users and OpenAI:
Allow users to submit product feedback directly through ChatGPT after reviewing and authorizing exactly what will be shared.
The Problem
I recently developed several detailed ChatGPT product recommendations through conversations with ChatGPT itself.
When I wanted to submit those ideas to OpenAI, I discovered that I still had to leave the normal conversational workflow, determine the correct feedback channel, manually copy information, search the Developer Community for existing requests, and submit the recommendations separately.
There seems to be an opportunity to make ChatGPT itself the interface for this process.
Proposed Workflow
A user could simply say:
“Submit this as product feedback.”
ChatGPT could then:
- Recognize that the user wants to provide product feedback.
- Summarize and structure the recommendation.
- Identify the relevant ChatGPT feature/product category.
- Show the user exactly what information would be submitted.
- Allow the user to edit or remove information before submission.
- Clearly indicate whether conversation context would be included.
- Require explicit user authorization before anything is transmitted.
- Submit the feedback through OpenAI’s appropriate internal feedback system.
- Provide confirmation that the feedback was received.
Nothing should be transmitted merely because ChatGPT believes something sounds like feedback. Submission should always require an explicit user action and final approval.
Privacy
Privacy should be built into the workflow.
Before submission, ChatGPT should clearly display:
- The feedback text
- Any conversation excerpts being included
- Whether account or diagnostic information is attached
- What information will not be included
The user could then approve, modify, or cancel the submission.
Duplicate Requests
An even more advanced version could recognize when similar feedback has already been submitted.
Instead of creating unnecessary duplicates, ChatGPT could potentially say:
“A similar feature has already been requested. Would you like to add your support and use case to the existing request?”
This could help OpenAI identify demand while reducing duplicate feedback.
Why This Could Benefit OpenAI
ChatGPT users frequently encounter product limitations while they are actively using ChatGPT.
At that moment, ChatGPT already has the context necessary to understand what the user was attempting to accomplish and why the limitation mattered.
A structured feedback workflow could potentially provide OpenAI with:
- Better-contextualized feedback
- More specific use cases
- Less fragmented feedback
- Better categorization
- Reduced duplicate requests
- Clearer indications of recurring user demand
Why This Could Benefit Users
The user would no longer need to figure out where feedback belongs or manually reconstruct an issue outside the conversation where it occurred.
The process could become:
Experience a limitation → discuss it with ChatGPT → develop a recommendation → review the submission → authorize it → submit.
Closing Thought
There is something particularly natural about using ChatGPT to improve ChatGPT.
If the assistant can already help a user identify a product problem, analyze possible solutions, and formulate a useful recommendation, allowing the user to explicitly authorize submission of that feedback seems like a logical final step.
ChatGPT could become not only the product being improved, but also one of the easiest interfaces through which users help improve it.