StableDiffusion was trained by an academic lab in Germany, which is rather different from a company in California doing it. Though they both have the problem that the output is illegal in many countries with the wrong prompts.