It's impossible to use AI to detect implied profane meaning, and this is not limited to French. You could slip your cucumber in your mother's oven and still not trigger any profanity detection software, as the implied meaning is only innuendo.
But it's useful to tease out obvious content in a fast efficient way.