and used to generate the answers. In one sense at least the digested data is searched to inform the answer. But it's not searching the original texts. I think there will be a very hard fought legal case, or cases, to determine if what I've been calling the "digested" data is, in copyright terms, 2/