All news

Risk prediction algorithms have become central to decision-making, despite known biases. Mitigation techniques too often focus on being statistically “fair,” overlooking the broader societal disparities these biases reflect. We need to expand the conversation to include the structural roots of bias for more transformative interventions.

The new Guide to Peer Review of Automated Decision-Making Systems published by the Treasury Board of Canada Secretariat (TBS) builds on the work of Dr. Kelly Bronson, Dr. Jason Millar and the team of the Canadian Robotics and Artificial Intelligence Ethical Design Lab (CRAiEDL) at the University of Ottawa's Centre for Law, Technology and Society.

Our critical systems are at risk of cyberattacks, especially the wireless networks that form the backbone of the 5G/6G era. With his expertise in AI-driven security solutions, Professor Burak Kantarci is tackling pressing security challenges of the digital era.

The rise of artificial intelligence is transforming public administration in Québec, particularly in the management of social benefits. The dematerialisation of social assistance files promises efficiency and agility, but also poses major challenges: the dehumanisation of service provision, reduced discretion for agents and the embodiment of a new form of algorithmic governance.

Equitable and accessible health data is essential for global health innovation. A collective approach and global frameworks would help share the benefits and unlock the exponential value-creation opportunities of artificial intelligence for health care.

The University of Ottawa AI + Society and IVADO release a global policy brief led by Prof. Florian Martin-Bariteau and Prof. Catherine Régis to provide policymakers with actionable globally-oriented policy guidance on the impact of AI on democracy and electoral integrity

The static nature of algorithmic impact assessments makes them ill-equipped to fully protect against embedded gender biases. A holistic approach that includes feminist methodologies and perspectives, like the “woman question”, is necessary to preemptively tackle these issues.

Current AI methods lack an understanding of legal decision-making. Legal meanings are often lost and expertise is at risk. We need new design frameworks to reintroduce lost meanings, prevent expertise loss, and regulate AI system design effectively.

Professor Karine Gentelet and students at the University of Ottawa have developed a reference directory on digital and artificial intelligence ethics for First Nations, produced with the support of the AI + Society Initiative.

Artificial intelligence (AI) is everywhere, including in the federal government. While AI can enhance efficiency, public servants must be mindful of its social and ethical challenges, including its potential for perpetuating biases.