For awhile, we have been trying to ‘get behind the curtains’ of what IRCC is doing on the AI decision-making front. We have a copy of a 2021 “Policy Playbook on Automated Support for Decision-Making Systems” but we have not really heard much over the past two years since then (and there has been no updated playbook).
Today we received slides that do give a bit of an update and summary of what has occured in the past two years. That being said, one should be mindful that the audience of this presentation is a Japanese delegate. This is not the PPT slides from an internal briefing where contentious issues are necessarily discussed. This is what IRCC wants the international public to know. From reading email correspondence and other internal briefings, we know there is a lot more tension in practice.
Below are a copy of the deck slides for everyone’s reference. Note the focus on the automated-decision making for overseas TRV use case:
Pages-from-A-2023-00081-Final-published-and-unpublished-Algorithmic-Impact-Assessments-performed-by-IRCCAny thoughts?
