Towards a Considered Use of AI Technologies in Government
The AI landscape is rapidly evolving and governments around the world are grappling with how to respond. The Government of Canada is starting to take steps to address governance of AI in the public sector. In this context, the IOG and Think Digital have recently released work examining how this powerful technology can be responsibly used and governed.
Our premise behind this report is simple: the AI genie is out of the bottle, and it is not going back in.
As a result, governments at all levels are increasingly going to have to grapple with how to responsibly use these powerful, compelling, and accessible technologies. Understanding what other jurisdictions are doing, what has worked, and what has not, is valuable for government officials as they consider the risks involved in using (or not using) AI in the public sector.
For this reason, the Institute on Governance and Think Digital undertook a case study-based research project, where 24 examples of AI technology projects and governance frameworks across a dozen jurisdictions were scanned. The purpose of this report is to provide policymakers and practitioners in government with an overview of controversial deployments of Artificial Intelligence (AI) technologies in the public sector, and to highlight some of the approaches being taken to govern the responsible use of these technologies in government.
Two environmental scans make up the majority of the report. The first scan presents relevant use cases of public sector applications of AI technologies and automation, with special attention given to controversial projects and program/policy failures. The second scan surveys existing governance frameworks employed by international organizations and governments around the world. Each scan is then analyzed to determine common themes across use cases and governance frameworks respectively. The final section of the report provides risk considerations related to the use of AI by public sector institutions across use cases.
This project was administered by the Institute on Governance and conducted by the team at Think Digital which included (in alphabetical order): Ryan Androsoff, Aislinn Bornais, Jacob Danto-Clancy, Bryce Edwards, Jen Schellinck, and John Stroud.
Interested in being part of the conversation on responsible use of AI in the public sector? Register now for our upcoming Governing AI in the Public Sector webinar, featuring paper authors and discussants. The webinar will take place on Friday, October 20, 2023 from 12-1 PM EST.