I think it very soon will be common practice to introduce a «security question» in types of conversations that could be fraudulent.