I start my analysis with the need to have good, clear, sufficiently high legal standards, including thresholds. That's where the issue of relevance for contributing to the mandate or being necessary comes in, so there are substantive legal safeguards.
The second element of well-balanced national security legislation requires strong, independent, effective review. On the substantive legal safeguards side, I accept that to apply the necessity test may pose problems for disclosing institutions, which is the main point the government made in responding to the ethics committee, and which may have been a contributing factor to your committee when you suggested a dual threshold.
I accept that a threshold lower than necessity helps disclosing institutions do a difficult task while having safeguards. However, receiving institutions—essentially national security agencies—know very well what their mandate is and what they need to do their job. There, the necessity threshold, which is the international norm, should apply fully.
That's the main substantive recommendation I'm making, which is again where this committee was at not long ago.
The second substantive rule is as follows. If there is a difference between the thresholds applicable to disclosing and receiving institutions which would be the result of a dual threshold, it's easier for disclosing institutions to disclose, but the threshold for receivers is higher. Point one is, what do we do about this gap, if the receiving institution has received something that is not necessary?
Point two is, if the receiving institution has received information about a law-abiding citizen—travellers are the best example—to identify in the mass of travellers the extremely few who may pose a threat to national security, there should be legal rules to require the receiving institution to get rid of the information, to destroy the information, to no longer retain the information if there's a gap between the two thresholds, or if, in relation to a given individual, the analysis leads to the conclusion that the person is not a threat and therefore that their information should not sit in the records of CSIS or the CSE or the intelligence apparatus. These are the substantive rules.
In terms of effective review, it is clear that the creation of the new NSIRA is an important improvement. The fact that it will be able to share information with the committee of parliamentarians creates a good step in the right direction, in that you have integrated review applicable to all departments—not only three as at the current time—and you have elected officials and experts who can talk to one another and reach a well-informed decision.
What we think we can bring to the picture—and we're not in the picture, at least not completely—with Bill C-59 is that the lifeblood, la matière première, the main tool that national security agencies have to do their job is information, and that includes personal information. We're the experts in how to deal with personal information in a way that respects privacy rights. We're not saying that NSIRA would be without any knowledge of the relevant issues, but there is an issue of core importance to the work of national security agencies, that of privacy, where we're the experts, and we think we can add value to the rest of the architecture.