Comments on “Guide on the use of Generative AI”

9 Pages Posted: 29 Nov 2023

See all articles by Blair Attard-Frost

Blair Attard-Frost

University of Toronto, Faculty of Information

Ana Brandusescu

McGill University

Renee Sieber

McGill University

Date Written: November 6, 2023

Abstract

On September 6, 2023, the Treasury Board of Canada Secretariat (TBS) published preliminary guidance on the use of generative AI (GAI) systems in Canadian federal institutions. The TBS Guide on the Use of Generative AI (hereafter referred to as the Guide) provides “an overview of generative AI, identifies challenges and concerns relating to its use, puts forward principles for using it responsibly, and offers policy considerations and best practices.” On October 13, 2023, we were invited by TBS staff to submit our feedback on the Guide. This document contains our comments on the Guide.

As a preliminary guidance document intended to provide provisional support to federal public servants amidst the rapid popularization of GAI applications, the TBS Guide is fit for purpose. Our comments identify three key opportunities for further strengthening the Guide and other related AI policy instruments:

(1) Integrate and build upon the TBS Guide, the TBS Directive on Automated Decision-making, and the Algorithmic Impact Assessment Tool to draft new federal legislation that will set more comprehensive and enforceable rules on the development and use of GAI and other AI systems in the federal public sector. This must also include the foundation models upon which fine-tuned models and applications are developed.

(2) Provide further guidance on ethical sourcing practices, particularly on ethical concerns related to copyright protection and worker protection in GAI value chains.

(3) Provide further guidance on practices for environmental impact mitigation.

Keywords: Artificial intelligence, generative AI, AI policy, AI governance, AI ethics

Suggested Citation

Attard-Frost, Blair and Brandusescu, Ana and Sieber, Renee, Comments on “Guide on the use of Generative AI” (November 6, 2023). Available at SSRN: https://ssrn.com/abstract=4624621 or http://dx.doi.org/10.2139/ssrn.4624621

Blair Attard-Frost (Contact Author)

University of Toronto, Faculty of Information ( email )

Ana Brandusescu

McGill University ( email )

Montréal, Quebec
Canada

Renee Sieber

McGill University ( email )

1001 Sherbrooke St. W
Montreal, Quebec H3A 1G5
Canada

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
222
Abstract Views
639
Rank
269,085
PlumX Metrics