Intégration de l'intelligence artificielle dans les processus de recrutement et sélection: quelle influence a le droit du travail et son principe de non-descrimination dans le processus de changement?

Abstract

The aim of this thesis was to answer the question "how do legal frameworks influence companies incorporating artificial intelligence into their recruitment and selection process?" and to see how companies managing their recruitment internally and recruitment agencies differed in their responses. We arrived at this question after realising that our literature review, starting with artificial intelligence, going through its use in recruitment and selection processes and the different legal frameworks of interest to algorithmic discrimination, had not informed us at all on this subject. We therefore conducted case studies with companies managing their recruitment internally, whatever their sector and size, and with recruitment firms, whatever their size. We met with fifteen companies for no less than eighteen people. We found that legal frameworks did not influence companies in their use of artificial intelligence, whether it was a company or a recruitment firm. We therefore found that the distinction between companies and recruitment firms is not justified. However, leaving aside legal frameworks, recruitment firms are more likely to use artificial intelligence in a more innovative way than companies. We pointed out that these results surprised us in two ways. Firstly, we did not anticipate that many firms would use external suppliers and therefore not be primarily concerned with legal frameworks. We tried to contact these external providers and the IT and legal departments of companies with their own tools, but they did not wish to participate in our study for reasons of confidentiality and lack of time. Secondly, we studied the anti- discrimination laws at length because we really thought that companies would report being much more concerned by them than they turned out to be. We concluded by saying that the same study in other countries more developed in artificial intelligence and in sectors more challenged by anti-discrimination laws would be relevant.