I'm trying to be as civil as I can, but apps like this should be treated as literal scams. AI detection doesn't and cannot work unless it looks for intentional watermarks left by llms. You are preying on people without domain knowledge.
On the off-chance this one does something different, I tested it on several pre-ai pieces of text - they were all marked as AI, including the declaration of independence which scored 91.35% AI-generated.