Assume for the sake of argument, then, that we recommend putting in a requirement for written agreements and we suggest to institutions that they have safeguards put in place for the reliability of information. Imagine that an individual works for Immigration Canada. They see a document that, in their view, has national security implications, but they are unclear on its full implications. They then would send it to CSIS, and they wouldn't be worried about doing so, because they're disclosing on the relevance standard, but CSIS, to actually collect that information and to use that information, is still subject to the “strictly necessary” test and any warrant requirements, as I understand it.
I understand that there's been confusion about the law, but if we put it in black and white that this is the case, what is the concern, then?