Safeguarding AI Use Through Human-Centric Design
Date: May 15, 2020
This paper examines the risks and ethical challenges that need to be addressed and overseen to ensure that the use of artificial intelligence (AI) provides benefits and avoids harm. The paper delves into practical and technical challenges and solutions, by bringing the perspectives of practitioners, academics, and technologists in Canada’s AI ecosystem (listed in the paper) on how some of the human-rights risks introduced by AI systems are being and can be managed during the product development life cycle.
Lead author for this paper was Donna Bales with contributions from Pavel Abdur-Rahman, Head of Trusted Data & AI, IBM Canada, Paul Finlay, Machine Learning Lead, silverhammer.ai, Matt Fowler, Vice-President, Enterprise Machine Learning, TD Bank Group, Jean-François Isabelle, Principal, Résolution Consulting Inc., Rheia Khalaf, M.Sc., FSA, FCIA, CERA, Director, Collaborative Research & Partnerships, Fin-ML; and member of the Partnerships team, IVADO, Manuel Morales, Associate Professor of Mathematics & Statistics, University of Montreal; Chief AI Scientist, National Bank of Canada; and Director, Fin-ML, Laila Paszti, Technology and Privacy Lawyer, Norton Rose Fulbright, Dominique Payette, Lawyer, Legal Affairs, National Bank of Canada, and collaborator, Fin-ML.