The DRIAS conference will discuss the ethical, legal and technical responsibilities of AI tools in health

7 juin 2021

June 21 and 22 will take place the DRIAS (Law, AI and Health) Colloquium, an initiative co-organized by the Institute for Research in Computer Science and Random Systems (IRISA). This event will be held online to discuss the technical, ethical and legal responsibilities in the use of AI models in the health field. This webinar will bring together about twenty specialists in AI, law and health.

A conference to better understand the ethical, legal and technical issues of AI tools in health

The DRIAS conference will allow to debate on the ethical issues and the regulation of artificial intelligence tools so that the rights of patients, as well as those of health professionals, can be respected. The notion of responsibility in case of exposure to risks related to the exploitation of AI will be highlighted. This is a concept that lies at the intersection of three disciplines and multiple questions:

Technical issues: to the designers of AI tools who have to think about the tools to meet the needs of identifying responsibilities (interpretability of decisions, certification of algorithms, protection of personal health data).
Ethical issues: to caregivers and decision-makers faced with contradictory injunctions between the expectation of benefits and the exposure to risks.
Legal issues: positive law will most probably have to be adapted to the standards being developed at the level of European Union law.

These three fields will be evoked during the DRIAS conference, which aims to offer Internet users crossed views of researchers in law, AI and the health world. Several sessions will be organized around the following themes:

Evolution of the European Union regulatory framework on AI tools in health;
Interpretability of decisions provided by AI tools, liability and accountability;
AI Certification;
Exploitation/protection of personal health data trade-offs:
Rights and obligations of healthcare professionals and patients in the face of the deployment of AI tools in health.

A program that promotes transdisciplinarity

Below, find the whole program of the DRIAS conference:

Monday 21 June 2021

Introduction (9h-9h15)
Roundtable: “Union law / AI / Health” (9:15-10:30) with:
- Brunessen Bertrand (Univ. Rennes 1/IODE), public law
- and Juliette Sénéchal (University of Lille), private law
Session “Issue of AI interpretability for liability accountability” (11am-12.15pm) with:
- Élisa Fromont (Univ. Rennes 1/IRISA/IUF), AI and interpretability
- Laurène Mazeau (UBO): “Legal accountability and explicability of decision support processes”)
- and Gérard Le Goff (France Asso Santé)
Session “Product defect / certification: which regime for AIs” (2pm-4pm) with:
- Hélène Muscat / Christine Paillard (University of Rennes 1/IODE/DRIAS): “Introduction to product liability”,
- Julien Chiaroni, pilot of the Innovation Council’s Grand Challenge on security, “The reliability and certification of systems that use artificial intelligence”,
- Guillaume Avrin (LNE), Head of the Artificial Intelligence Assessment Department
- and Pierre Saurel (Centre Boreli/DRIAS) “FDA certification of medical devices integrating AI”.

Tuesday 22nd June 2021

Introduction (9h-9h15)
Session “Rights and obligations of healthcare professionals and patients in the face of AI deployment in healthcare” (9:15-11:00) with:
- Anne Strauss: History and example of AI methods in health,
- Prof. Stéphane Oustric (CNOM), general delegate for data and digital at the National Council of the Medical Order,
- and Jean-Baptiste Thibert (Établissement français du Sang), Physician/Lawyer, consent
Session “Rights and obligations in the design and use of AI systems in Health” (11:15-12:00) with:
- Pierre Saurel, Danièle Bourcier, “The right not to be subject to a decision based solely on automated processing (Art. 22 RGPD)”
- and Jennifer Verney Dahan, “Legal design”
Session “Exploitation of personal data in Health” (2pm-4pm) with:
- Sébastien Gambs (UQAM) “anonymization and pseudonymization techniques”,
- Félicien Vallet (CNIL), Anonymisation,
- Pierre-Alain Jachiet (HAS, Head of data strategy)
- and Étienne Audureau (AP-HP, Chaire AIRACLES), “Exploitation of data from the AP-HP health data warehouse