News

This page contains automatically translated content.

Back
01/06/2021 | Gender/diversity in informatics systems

EU-Canada Cooperation Workshop in AI: Equity, Diversity and Inclusion in Artificial Intelligence

The workshop will take place on April 26 and 27, 2021.

The aim of this workshop activity is to identify potential collaboration priorities in research in equity, diversity and inclusion in AI under the upcoming Horizon Europe Programme (HEU 2021-2027), in order to boost EU-Canada STI collaboration activities in the area. The meeting will consist of a day for policy discussion and identification of common priorities and foresighting, and a second day for the identification of potential ways forward and mechanisms for cooperation by funding decision-makers (joint programming, twinning, program alignment etc.).

1. Context and Approach

The European Commission funded the creation of an EU-Canada Programme Level Cooperation Task Force, in the frame of the International Service Facility of the DG Research and Innovation. The objective of the Task Force is to advance programme level cooperation between the European Union and Canada across the science, technology and innovation programmes in order to support and encourage their alignment and cooperation in the most effective and efficient possible way.
The Task Force is co-chaired by the European Commission and Global Affairs Canada. From the Canadian side, the Members are representatives of the three federal granting councils (CIHR, NSERC, SSHRC); Canadian Funding Agencies as CFI, and NRC; Ministries as AGRI Canada, DFO (Department of Fisheries and Oceans and Environment, etc.), plus a good representation of Canadian Provinces and Territories (the most active ones are Québec, Ontario, British Columbia and Alberta).
The workshop on Artificial Intelligence is one of the four thematic workshops aimed at exploring new areas of potential cooperation between EU and Canada.

2. Objectives

The aim of this workshop activity is to identify potential collaboration priorities in research in equity, diversity and inclusion in AI under the upcoming Horizon Europe Programme (2021-2027), in order to boost EU-Canada STI collaboration activities in this theme.
The workshop participants will have the chance to:

  • Share information on related priorities and flagship projects from EU and Canada
  • Discuss and recommend potential collaboration topics - and relevant impact
  • Discuss action points – short term and longer term goals
  • Discuss framework conditions and ways to collaborate

3. Theme

The workshop’s theme addresses issues of equity, diversity and inclusion in AI’s development and application. Specific factors would include gender, race, ethnicity, age, Indigenous or Aboriginal identity, and identity as a person with a disability.
In the context of machine learning, bias can mean that there is a greater level of error for certain demographic categories. Because there is no one root cause of this type of bias, there are numerous variables that researchers must take into account when developing and training machine-learning
models. Gender bias has been at the center of the attention, however recently the importance of considering the bias around all under-represented groups has been increasing.
There is a diversity crisis in the AI sector including gender, visible minorities, Indigenous and Aboriginal people, persons with disabilities, and age. The AI sector needs a profound shift in how it addresses the current crisis to create fully robust AI for the 21st century. Currently, many sectors of society are left out of the creation and development of AI, which when implemented does not capture their experiences or realities despite the fact that it is already being used to make decisions about their lives. As a result, many people are being excluded and disempowered and the resulting technology is biased.
New responses are required at the global level as AI does not respect political boundaries. As studies have shown in relation to increasing the number of women in STEM, fixing the “pipeline problem will not fix AI’s diversity problems which stem from systematic biases”. This workshop will allow Canada and the EU to explore what aspects of this issue may be possible to address through future collaboration.