Colloquium | Hybrid
As part of the 91st ACFAS convention
Artificial intelligence (AI) is making inroads into every facet of social life: every day brings its share of examples of AI use, whether positive or negative for society and individuals. The emergence of generative AI, for example, is shaking up the world of culture and challenging the business model. Healthcare uses are promising, but the benefits that can be derived from them still need to benefit everyone. Whatever the case may be, it's clear that AI technologies generate risks not only of breaching confidentiality and privacy, but also of creating biases that can lead to discrimination. As a result, there is a growing recognition of the need for ethical, technical and, above all, legal solutions to regulate the use of AI.
Now that Canada and the European Union have presented their respective draft laws on artificial intelligence, many questions remain: Are the proposed rules appropriate? Are they sufficiently precise and effective? Are there not risks of rapid obsolescence? Are technical specificities properly taken into account to ensure that the subject matter is properly captured?
This analysis follows a comparative law methodology and a multidisciplinary approach, integrating ethics, law and AI sciences (computer science and mathematics). More specifically, ethicists, jurists and scientists examine the definitions and regime of AI, in particular foundation models such as generative AI. They deserve to be the subject of a multidisciplinary dialogue in order to fully grasp their meaning and scope, and to assess the need for special rules.
The critical and comparative evaluation of draft legislation must also question the relevance of alternative solutions.
