Two McGill professors awarded funding for AI research through Canada-France collaboration

McGill Professors Jackie Chi Kit Cheung and Lijun Sun have received funding for their research projects on reasoning and responsible artificial intelligence (AI) from the Canadian Institutes of Health Research (CIHR), Natural Sciences and Engineering Research Council of Canada (NSERC), and Social Sciences and Humanities Research Council (SSHRC), announced on November 17.  

The funding comes from the Canada-France call for proposals on AI, a joint research initiative launched in 2024 by the Agence nationale de la recherche (ANR) in France and Canada’s Tri-agencey partners—NSERC, SSHRC, and CIHR—with support from the Institut de valorisation des données (IVADO).  

The program supports collaborative, multidisciplinary projects from within the Canadian and French research ecosystems focused on generative AI and the security and safety of embedded AI. Ten projects have received a total of $3,292,500 in funding in this competition.  
 

Recipients: 

Jackie Chi Kit Cheung, Associate Professor, School of Computer Science 
Project: REGARD - Robust Evaluation of Generative AI in Real-World Environments 
Collaborators: Edith Law (Waterloo University), Pablo Piantanida (Centrale Supélec), Frédéric Béchet (Université Aix-Marseille), and Geraldine Damnati (Orange Labs). 
Funder: SSHRC ($300,000) and IVADO ($60,000) 
The REGARD project aims to advance the evaluation of language generation systems by understanding their behaviour, measuring performance, generalization, and robustness, and assessing their societal impacts. By integrating social, algorithmic, and theoretical perspectives, Cheung will develop frameworks, protocols, evaluation methods, and case studies to analyze benefits and harms, assess limitations, and ensure models are reliable, value-aligned, and applicable in real-world settings. 

Lijun Sun, Associate Professor, Department of Civil Engineering  
Project: MobAgent - Large Language Model Agent for Mobility Reasoning and Synthesis 
Collaborator: Latifa Oukhellou (Université Gustave Eiffel) 
Funder: NSERC ($300,000)
MobAgent will create realistic, high-resolution human mobility data from diverse and sparse sources like transit cards, call records, and social media—sources that current AI methods struggle to integrate, thereby limiting their application in transportation planning and urban management. Leveraging large language models to enhance deep generative models,  MobAgent will produce privacy-preserving data that can capture complex travel patterns. These methods will provide reproducible, adaptable mobility data to support sustainable transportation, equitable urban access, and evidence-based planning, while fostering international France-Canada collaboration. 

Photograph: McGill Professors Jackie Chi Kit Cheung and Lijun Sun