Workshop : Which paths to achieve fairness in algorithmic decisions? Replay
Le workshop Which paths to achieve fairness in algorithmic decisions? qui s’est tenu en ligne les 9 et 10 décembre 2021 a été organisé par les centres de recherches dauphinois : l’IRISSO, le LAMSADE et le CR2D dans le cadre du projet INTER-FAIR soutenu par la Mission pour l’Interdisciplinarité et les Initiatives Transverses du CNRS en 2020-2021 : La décision algorithmique : comment penser l’intégration de l’éthique comme valeur sociale? Approches interdisciplinaires.
Le workshop, au croisement des SHS, des Data Science et de l’informatique, portait sur des thématiques également privilégiées par le Programme gradué Data Science.
Le comité d’organisation était composé de : Jamal Atif (LAMSADE), Thierry Kirat (IRISSO), Antoine Louvaris (CR2D), Olivia Tambou (CR2D), Virginie Do (LAMSADE) et Céline Beji (LAMSADE).
Session 1
International workshop Which paths to achieve fairness in algorithmic decisions? Université Paris Dauphine-PSL, 9-10 décembre 2021
Session 1 - Legal Governance of Algorithms: hard law or soft law? (9 décembre 2021)
- Raja CHATILA (Sorbonne Université): The Path to Trustworthy AI Systems: Principles, Requirements, Standards and Governance
- Karine GENTELET (Abeona-ENS-OBVIA Chair & Université du Québec en Ouataouais): Human Rights: Pilar of AI Regulation(s)
Session 2
International workshop Which paths to achieve fairness in algorithmic decisions? Université Paris Dauphine-PSL, 9-10 décembre 2021
Session 2 - Metrics of fairness : legal and/or technical issue ? (9 décembre 2021)
- Alice XIANG (AI Ethics Lead, Sony AI): Reconciling Legal and Technical Approaches to Algorithmic Bias
- Solon BAROCAS (NYC lab of Microsoft Research & Cornell University): Explanations in Whose Interests?
- Aurélien GARIVIER (Ecole Normale Supérieure de Lyon, UMPA): How to include fairness into the machine learning models?
Session 3
International workshop Which paths to achieve fairness in algorithmic decisions? Université Paris Dauphine-PSL, 9-10 décembre 2021
Session 3 - Explicability, transparency, accountability : constraints and possibilities for public policy (10 décembre 2021)
- Doaa Abu ELYOUNES (Harvard Law School, Berkman Klein Center for Internet and Society): “Computer says no!”: The Impact of Automation on the Discretionary Power of Public Officers
- Lilian EDWARDS (Newcastle University Law School): Getting a Global Regulatory Model for AI Right
- Thierry TUOT (Conseil d’Etat): From Lost illusions to Time regained : in search of real user’s rights
Workshop presentation
This workshop address possible current and future paths to achieve fairness in algorithmic decisions, both private and public.
International and multidisciplinary, the workshop focus on the intersection and cross-fertilization of computer science, law, and policy analysis. It addresses the question of how to regulate algorithmic decisions, through hard law or soft law. It also addresses the possibilities and obstacles to complementarity between computer science and legal approaches to fairness.
Finally, the workshop aims to shed light on the policy issues surrounding explicable, transparent and accountable algorithms.
