I would start first with where there was agreement beyond the need for the law to be changed.
Many people felt that the guidance was drafted or crafted at a level of generality such that the advice is helpful, but they would like it at the very least to be supplemented by advice on what was called “use cases”. Our reaction to that is that indeed there is a need for advice on particular uses in different contexts, because context matters a whole lot, but we still think it's important and relevant to have general guidance that can be be augmented as use cases are developed.
Some stakeholders from civil society or minority groups called for a moratorium on the use of facial recognition. The RCMP obviously did not agree with that. Our position as commissioners is that there should be clear laws prescribing when facial recognition can be used, because it can be used for legitimate, helpful purposes and social good in some circumstances—for instance, in serious crime situations or to find missing children—but these uses should be defined quite narrowly. The law should also prescribe prohibited uses, which would be, I guess, a partial ban or a partial moratorium on the use of facial recognition.
If I may, on the question of a moratorium, we as data protection authorities cannot impose a moratorium that has the force of law. For a moratorium to be binding on police agencies, it would have to take the form of legislation.
I was struck by the testimony that you heard last week from an RCMP representative, to the effect that “The RCMP believes that the use of facial recognition must be targeted, time-limited and subject to verification by trained experts.”