New content filter preferences in playground

You can now control whether warnings about sensitive or unsafe content appear when generating completions in the playground.

In the playground sidebar you’ll now see a link to manage your content filter preferences:


This will open a dialog that allows you to toggle warnings for sensitive and unsafe completions:

Note that the content filter will still run in the background for safety purposes, you will just be able to disable the warnings if you find that they’re disrupting your experience. We will also continue to block prompts that have a high risk of producing unsafe content.


The changes outlined here only affect the playground, the API itself is unchanged.