On March 12, 2026, the Senate Standing Committee on Social Affairs, Science and Technology invited Tania Saba, Interim Executive Director of Obvia, Full Professor at the School of Industrial Relations at Université de Montréal, and holder of the BMO Chair in Diversity and Governance, to appear as part of its study on the impacts of artificial intelligence (AI) in Canada.
This initiative is part of a broader reflection examining several key issues, including AI governance, data sovereignty, ethics, privacy protection, as well as the societal risks and impacts associated with the deployment of AI.
While Canada benefits from internationally recognized scientific leadership in AI, adoption remains limited, particularly among small and medium-sized enterprises (SMEs). Key barriers include limited access to financial and technological resources, a lack of internal expertise, and restricted access to the data required for the development and deployment of AI systems.
“Canada has very strong scientific leadership in AI, and the pressure to adopt these technologies is significant. However, adoption is not happening at the expected pace, particularly among SMEs, where only 12% of businesses facing specific barriers are adopting AI.”
-Tania Saba, Interim Executive Director of Obvia
During her appearance, Tania Saba reiterated Obvia’s essential role in advancing independent research on the societal impacts of artificial intelligence and digital technologies, while highlighting its interdisciplinary and intersectoral approach rooted in the social sciences. Her remarks then focused on three fundamental questions related to AI governance.
In the face of rapid transformations driven by artificial intelligence, what ethical principles should guide AI development?
Over the past several years, several international organizations have proposed comprehensive ethical frameworks, including UNESCO, the OECD and various initiatives emerging from international forums.
These frameworks reveal a broad convergence around fundamental principles: human-centred AI that respects fundamental rights and democratic values; the need to ensure transparency, explainability and accountability of algorithmic systems; the importance of guaranteeing the safety, robustness and reliability of technologies; as well as the promotion of responsible innovation attentive to economic, social and environmental impacts.
How can these principles be strengthened in a fragmented and competitive environment?
While governments play a central role, AI governance now relies on a broader ecosystem that includes scientific communities, research centres, international organizations, technology companies and international collaboration networks.
These collaborations bring scientific research, public policy and societal concerns closer together, contributing to the emergence of a form of scientific diplomacy. However, several challenges remain: coordination among stakeholders, diverging priorities between governments, businesses and scientific communities, tensions between economic interests and regulatory objectives, as well as the fragmentation of international cooperation mechanisms.
One of the major challenges for the coming years will be to strengthen spaces for dialogue and cooperation among these different stakeholders in order to better align technological innovation, scientific expertise and public governance.
-Tania Saba, Interim Executive Director of Obvia
How can ethical principles be translated into concrete mechanisms?
It is essential to move beyond a purely declarative approach to ethics and towards operational AI governance. This requires, in particular, the development of scientific indicators and tools capable of providing concrete guidance for the development and use of artificial intelligence systems.
This shift is reflected in the emergence of normative and regulatory frameworks designed to operationalize ethical principles through risk management standards, data quality requirements, transparency measures and system oversight mechanisms. At the same time, disclosure practices and organizational governance are gradually incorporating AI-related considerations, particularly regarding risk management and accountability.
In this context, the development of measurable indicators becomes essential to evaluate system performance, document their impacts on organizations and society, and strengthen accountability.
However, challenges remain, including fragmented standards, limited data on the real-world impacts of AI, and the difficulty for SMEs to understand and integrate these frameworks into their practices.
Ultimately, the challenge is no longer only to define ethical principles for AI, but to translate them into concrete mechanisms capable of guiding development and adoption practices, measuring technological impacts, ensuring accountability — which is essential to addressing the burden of proof — and supporting the development of AI that is truly responsible and sustainable.
