AI Governance on the Ground Series

The minute decisions, measurements and methods embedded inside the tools used to govern AI systems directly affect whether policy implementations actually align with policy goals. The government of Chile’s experience using its AI bidding template, and questions inside the agency regarding how to weigh traditional tech procurement criteria such as vendor cost along with newer responsible AI criteria like discriminatory impacts, give a glimpse of the AI governance challenges happening on the ground today. The tensions the Chilean government is dealing with may be a sign of what other organizations around the world could encounter as they put their own responsible AI policies into practice and navigate the policy implications of AI-facilitated decision making.

It is my great pleasure to introduce our new podcast, Privacy on the Ground. This podcast is about all of the things World Privacy Forum has been researching, writing about and working on for more than 20 years, but in a format for today’s audiences. Our groundbreaking work on data governance, data protection, privacy, and complex technologies and ecosystems is often concentrated in dense reports that can take years to research and produce. Our podcasts enable us to bring the compelling people, stories, places, news, and ideas we find in the course of our work to the public faster and in an episodic format that allows us to highlight the richness and depth of the people, stories, trends, and ideas you won’t find anywhere else.

WPF’s “AI Governance on the Ground Series” highlights and expands on topics and issues from WPF’s Risky Analysis report and its survey of AI tools. In this first publication of the series, we highlight how Canadian government agencies are implementing AI governance and algorithmic transparency mechanisms across various agencies, including its employment and transportation agencies, its Department of Veterans Affairs, and the Royal Canadian Mounted Police, among others. The agencies have evaluated the automated systems they use according to the country’s Algorithmic Impact Assessment process, or AIA, and the assessment results are public. Designers of this assessment framework — required since the country’s Directive on Automated Decision-Making went into effect in April 2019 – have now re-evaluated the AIA, updating its criteria, requirements, and risk-level scoring algorithm along the way. WPF interviewed government officials as well as key Canadian end-users of the assessments to capture the full spectrum of how the AIA is working at the ground level.