Microsoft expands Azure OpenAI Service with DALL-E 2 in preview

  To prevent misuse, as with Designer and Image Creator, Microsoft says it’s implemented filters to reject DALL-E 2 prompts from Azure OpenAI Service customers that violate content policy. The company also claims it’s integrated techniques to prevent DALL-E 2 from creating images of religious objects and celebrities, plus objects commonly used to try to trick the system into generating sexual or violent content. And Microsoft says it’s added models that remove AI-generated images appearing to contain adult, gore and other types of “inappropriate” content.

And this is why I would never sign up for something like this. I've been using stable diffusion a bit, for a completely benign commercial / research application that has nothing to do with anything related to people or remotely NSFW. When I first started using it I was regularly triggering the NSFW filter, with false positives, that after removing the filter I confirmed were sfw and looked like the rest or what I was generating. Adults don't want or need to have their hands held because of some SV activists opinion about what "harms" we might encounter. That kind of filtering has no place in a tool like this