I’m submitting multiple images to GPT-4V with the same prompt and I get
“Your input image may contain content that is not allowed by our safety system.”.
Upon manual inspection, some images have mummies so I get why the system might flag them as no safe. However, the majority of images are alright.
I do need to submit all the images in one go because the task is to compare each image. However, I would expect the system to tell me exactly what image has the issue so I can exclude it and run the prompt again. I can’t do anything with “Your input may contain…”.
Another alternative might be to automatically ignore images with issues and raise a warning.