Like an ouroboros—or snake eating its own tail—future AI models trained on the internet, where AI-generated content is expected to become ubiquitous, could end up devouring the problematic work of their predecessors.

The next generation of state-of-the-art foundation models could have capabilities that severely threaten public safety and global security. In a new multi-author white paper, the authors discuss the challenges of regulating frontier AI models, explore the building blocks of a potential regulatory regime, and suggest some minimum safety standards.

A new paper by Aparna Balagopalan, David Madras, David H. Yang, Dylan Hadfield-Menell, Gillian Hadfield, and Marzyeh Ghassemi describes how decisions made by machine learning models can be improved by labelling data that explicitly reflects value judgments.

In a recent article in PLOS Digital Health, SRI Graduate Affiliate Vinyas Harish explores how cross-sector partnerships can enable new technologies for more effective healthcare during emergencies, and the importance of robust governance frameworks.

Can government regulation benefit cybersecurity in the private sector? What requirements are needed to better assess threats and reduce vulnerability for Canadians? To answer these questions, SRI Research Lead David Lie led a panel discussion on the future of cybersecurity regulation.

In his new book Power and Prediction: The Disruptive Economics of Artificial Intelligence, SRI Research Lead Avi Goldfarb makes a compelling case for how AI innovation will evolve, and why it will require systems-level changes to the ways public and private sector organizations make decisions.

How can computing researchers better consider the ethical and societal impacts of their work? In a recent SRI Seminar, Harvard professor Barbara Grosz outlined the key findings of a major new report on responsible computing research, which provides practical recommendations for new methodologies and approaches.

How is virtual reality changing the way we live? Does living in a virtual landscape create new ethical and moral problems? At the first annual Technophilosophy September Soirée, SRI Research Lead Karina Vold convened an interdisciplinary panel of University of Toronto scholars to discuss the impacts of new technologies on the way we understand our world.

SRI Graduate Fellow Reid McIlroy-Young’s new research explores the concept of “mimetic models”: algorithms designed to simulate the behaviour of an individual in new situations. In a new paper, McIlroy-Young reflects on the ethical implications of such models, in what scenarios they might be used, and directions for future research.

The Schwartz Reisman Institute for Technology and Society at the University of Toronto has appointed five new research leads to help shape the Institute’s ongoing research program. The new research leads bring expertise and insights on the social impacts of artificial intelligence from the fields of economics, computer science, literature, and philosophy.

In a session at Absolutely Interdisciplinary 2022, SRI Associate Director Peter Loewen, Jonathan Stray of the Berkeley Center for Human-Compatible AI, and Taylor Owen of McGill University discussed what methods and principles might be used to redesign the algorithms that decide what billions of people see in accordance with democratic values.

The rise of the information society and the ubiquity of digital technologies brings new challenges for governance, as traditional institutions might not be sufficient to address the complex issues emerging from these new tools. In a talk at Absolutely Interdisciplinary 2022, SRI Faculty Fellow Anna Su explored the phenomenon of digital constitutionalism and the importance of human rights to balance powers in the digital realm.