Privacy policy of recent gpt-5 prompt optimizer tool

API
, , , ,
1

I’m using the recent prompt optimizer tool release around 2.5 months ago:

https://platform.openai.com/chat/edit?optimize=true

Does someone knows if this tool complies with the same privacy policy for API?

From this help article:

By default, we don’t use any inputs or outputs from our products for business users, including ChatGPT Business, ChatGPT Enterprise, and the API, to improve our models.

Looks like inside openai platform webapp there is a specific setting to control this:

https://platform.openai.com/settings/organization/data-controls/sharing

Enable sharing of model feedback from the Platform

Enable all members of your organization to share feedback and chats with OpenAI to help improve and train our models. If enabled, users can click the “thumbs down” button on model responses in the Playground and choose to share their feedback and content from the current chat.

Share evaluation and fine-tuning data with OpenAI

Turn on sharing with OpenAI for all evaluation and fine-tuning data from your organization, including evaluation and fine tuning prompts, completions, grading logic, and results, to improve and train our models. Only evaluation and fine-tuning data created after turning this setting on will be shared. You can change your settings at any time to disable sharing evaluation and fine-tuning data.

Share inputs and outputs with OpenAI

Turn on sharing with OpenAI for inputs and outputs from your organization to help us develop and improve our services, including for improving and training our models. Only traffic sent after turning this setting on will be shared. You can change your settings at any time to disable sharing inputs and outputs.

Setting all of them to Disabled (or at least the last one - Share inputs and outputs with OpenAI - just is more directly related to inputs and outputs) will ensure prompts shared in recent gpt-5 based prompt optimizer tool won’t be used for training? :thinking:

Because after exploring the API call with this recent tool:

https://api.openai.com/v1/dashapi/optimize/promptv2

{
  "developer_message": "<pre-optimized prompt>",
  "messages": [
    {
      "role": "user",
      "content": ""
    }
  ],
  "model_name": "gpt-5",
  "requested_changes": "",
  "tools": [],
  "optimizing_for": "gpt-5"
}

despite the fact of having api subdomain, I’m not sure if is following same policies as:

https://api.openai.com/v1/responses

If there is any explicit mention about this topic in:

  • policies url
  • tweet from some account related to openai (like this one)
  • or any other trustable, reliable, canonical source

I’d appreciate a lot if could be shared.

Thanks :folded_hands: