ED-AIM (Ethical Design for Artificial Intelligence Models in patient management and treatment decision)

As part of the Prime80 program of the Mission for Transversal and Interdisciplinary Initiatives (MITI in French) of the French National Center for Scientific Research (CNRS)

Presentation

ED-AIM stands for “Ethical Design for Artificial Intelligence Models in patient management and treatment decisions”. Funded by the Prime80 program of the Mission for Transversal and Interdisciplinary Initiatives (MITI in French) of the French National Center for Scientific Research (CNRS), ED-AIM is an interdisciplinary collaboration between the Maison Française d’Oxford (MFO), the Institute for the History of Representations and Ideas in Modernity (IHRIM, UMR 5317, ENS Lyon), the Research Institute in Computer Science and Random Systems (IRISA, UMR 6074, Rennes 1), the National Institute for Research in Digital Science and Technology (INRIA—Lyon antenna) and the Institute of Biomedical Engineering at Oxford University. The project is led by Thomas Guyet (INRIA), Mogens Lærke (MFO, IHRIM), and Pascal Marty (MFO).

The project aims to study the deployment of Artificial Intelligence (AI) models and systems in healthcare settings from a philosophical perspective. This latter perspective is double, both epistemological and ethical. On the one hand, understanding the epistemology of AI is required for critically assessing the potential changes it can bring to medicine. On the other hand, the ethical stakes surrounding AI in healthcare are multiple and draw on views from the philosophy of medicine and biology and their considerations on what it means to treat a patient ethically.

In ED-AIM, we explore not only the ethical expectations of AI-based systems themselves but also the socio-technical context in which such systems are to be deployed. We do this according to two main axes: defining what we can call ethical AI systems and models and devising ways to design such ethical systems. In a nutshell, the ambition is to address how to ethically design ethical AI-based systems in medicine with ontological, axiological, deontological and practical answers.

In practical terms, the ED-AIM project has two main components, namely a PhD funding (2021–2024) in philosophy for Éric Pardoux, and the upcoming organisation of events linked to the ethical stakes of the introduction of AI in medicine.

 

Publications

Peer reviewed articles and proceedings:
•    Pardoux, Éric and Louis Devillaine, “Vers une éthique processuelle de l’IA”, 2022, PFIA, Association Française pour l’Inteligence Artificielle, Saint-Étienne (Fr), https://hal.archives-ouvertes.fr/hal-03866133
•    Pardoux, Éric, "Ethical Design for AI in Medicine", 2022, Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society, page 907, https://doi.org/10.1145/3514094.3539564
•    Pardoux, Éric and Thomas Guyet, "Pro-ethical platforming of epidemiological modelling", pending.

Non peer-reviewed (public engagement):
•    Pardoux, Éric and Mogens Lærke and Thomas Guyet, “Enjeux éthiques de l’usage de l’intelligence artificielle en médecine : apports d’un regard épistémologique”, 2022, Association Française pour l’Intelligence Artificielle (Ed.), Bulletin de l’AFIA , No. 116, p. 41-43, https://afia.asso.fr/wp-content/uploads/2022/05/116_avr22.pdf

Talks

•    Pardoux, Éric, “The ethics of AI ethics (in healthcare) or “How to ethically design ethical AI?””, 21/02/2023, Workshop Ethical Design for AI in Medicine, Maison Française d’Oxford, Oxford (UK).
•    Pardoux, Éric, “Where to put ethics in AI?”, 08/02/2023, Internal seminar, Ethox Centre, Oxford (UK).
•    Pardoux, Éric, “Quand l’IA fait système – Exploration des dimensions sociotechniques et relationnelles de l’intelligence artificielle”, 12/12/2022, Séminaire Les concepts de la computer ethics, Chaire Éthique&IA de Université Grenoble-Alpes, Grenoble (Fr)
•    Pardoux, Éric, “Where to put philosophy in AI?”, 29/11/2022, Workshop Challenges of AI in biology and medicine, PhilInBioMed laboratory, Bordeaux (Fr).
•    Pardoux, Éric, “Ethical Design for AI in Medicine”, Student Program AIES 2022, Oxford (UK), 01 to 03/08/2022.
•    Pardoux, Éric and Louis Devillaine, “Towards processual AI ethics” , 2022, “Where AI ethics should go?” workshop, Tübingen (De), 30/06/2022 and 01/07/2022.
•    Pardoux, Éric and Devillaine, Louis “Vers une éthique processuelle de l’IA” , 2022, Plate-Forme Intelligence Artificielle, Saint-Étienne (Fr), 27/06/2022. 
•    Pardoux, Éric, “Usages de l’intelligence artificielle en santé — Questions éthiques et apports d’un regard épistémologique”, 12/12/2021, Réseau Idées sur la santé, Lyon (Fr).

Members

Researchers contributing to the ED-AIM project: