It's true however I do feel that when talking about scams AI can be used offensively way easier. Using it for things like the prevention of phone scams seems quite a bit more difficult to me to get accurate.