Start – 2018 | End – 2024 Duration – 6 years
Several abstract models have been designed to support the development of decision-making instruments. Both their experimental implementation and their practical evaluation are hindered by the reluctance on the part of those who render legal decisions. Transparency and pedagogy efforts are necessary to ensure the relevance and reliability of these tools, which implies examining already existing decision-support techniques.

Subproject chief
Fredric Lederer
Research activities
Case studies
AI prediction solutions used in the field (and preferably with a widespread use) will be addressed in case studies. Currently, at least one of them, COMPAS, is being used in the United States for pretrial release and criminal sentencing, and merits a case study.
Inventories
The objective is to determine the extent to which AI predictions are now being used in legal professions. These predictions are either under development or under serious consideration. Initial research has identified 93 potential “solutions” designed to assist legal professionals.
Best Practices Guide
The subproject group will work towards producing a best practice guide designed to bolster reliable results while protecting legal rights and minimizing legal liability.
Governance Framework
The subproject group will explore the creation and application of general principles to be used in determining whether and how governmental organizations, including courts and adjudicative agencies, should use AI prediction.
Researchers publications :
- Fredric LEDERER, "Here There Be Dragons – The Likely Interaction of Judges with the Artificial Intelligence Ecosystem", (2020) Vol 59 - No 1 The Judges Journal ; American Bar Association | Judicial Division. HTML
Partners
Institutional partner |
Institutional partner |
Institutional partner |
Institutional partner |
Institutional partner |
Industrial partner |
Industrial partner |
Professional partner |
This content has been updated on 3 September 2020 at 14 h 49 min.