Start – 2019 | End – 2024 Duration – 5 years
Data collected in the judicial field are sensitive and often need to be anonymized. Access to this data must be subject to regulatory mechanisms that clarify what must be deleted, what can be used and exchanged, under what conditions and for what purposes.
A first case study to test what type of data can be extracted from the electronic records and if AI tools can be used to generate new information about individuals from this data will be carried out.
A second case study will focus on identifying core privacy concerns and issues associated with AI tools and to elaborate solutions to alleviate these concerns.
The subproject will aim to do two inventories of studies, guidelines and best practices. The first will concern the technical and legal norms guiding online publication of court records.
The second will focus on privacy risks associated with AI tools in the context of conflict resolution and prevention, especially as they are relative to conversational agents.
A Best Practices Guide and Governance Framework for the publication of online court records will be produced.
The subproject will also make recommendations on privacy risk management for the use of AI tools in the context of ODRs.
Researchers publications :
- Ignacio COFONE, "AI and Judicial Decision-Making", in Florian MARTIN-BARITEAU, Teresa SCASSA (dir.), Artificial Intelligence and the Law in Canada, Toronto: LexisNexis Canada, (2021).
- Pierre-Luc DÉZIEL, Karim BENYEKHLEF, Eve GAUMOND (dir.), "Repenser la protection des renseignements personnels à la lumière des défis soulevés par l’IA" working paper, Montreal, OBVIA, April 2020. PDF
- Lyse LANGLOIS, Réjean ROY, Guillaume MACAUX et Eve GAUMOND (dir.), « Analyse sur l'application de notification de contacts proposée par Mila », working paper, OBVIA, June 2020. PDF
- Pierre-Luc DÉZIEL, "Petit guide sur les enjeux et opportunités des applications de notifications d’exposition à la COVID-19", OBVIA, June 2020. PDF
This content has been updated on 24 November 2020 at 12 h 52 min.