You may sometimes see the error “Your request activated the API's safety filters and could not be processed” on Board or "NSFW content detected" on Canvas.

Behind the scenes, we have an AI model that determines if a generated image is acceptable to return back to you. In most cases, this filter is accurate and won’t limit your generation of images, but sometimes the filter will be triggered even when it likely shouldn’t.

Our goal isn't to censor people unnecessarily but it's to limit malicious actors who might pose great harm to the world that we combat every day.

To lessen this issue, you can use negative prompts with words like "nudity, naked, nude, nsfw, pornography, porn, child, children":

Over time as we work on this, the filters will get much better.