Feedback for the OpenAI Development Team

Your image-generation safeguards need serious improvement.

A straightforward, non-explicit creative portrait request was rejected twice, despite using an established fictional character and previously accepted visual references.

Users shouldn’t have to fight the system to accomplish ordinary creative tasks.

Stop treating harmless requests as policy violations. Improve the accuracy, transparency, and consistency of your moderation systems.

Users deserve better than unexplained refusals and automated apologies.

Do better.