När får algoritmer övervaka i arbetslivet? - En rättslig analys av när algoritmisk övervakning och arbetsledning inskränker på arbetstagares integritet

Sammanfattning

This thesis examines when the use of algorithmic surveillance and management in the workplace becomes an unlawful intrusion into employees’ privacy. Due to rapid technological development, where digital tools and artificial intelligence are increasingly used to monitor, direct and evaluate work, the thesis aims to analyse this legal boundary under current law. The focus is on the requirements under EU law and Swedish law when using such systems and whether the current legal framework is sufficient to meet these developments. To answer these questions, the thesis uses a legal dogmatic method together with an EU law method. The analysis shows that there is no unified legal framework governing algorithmic surveillance and management. Instead, the legal position is shaped through an interaction between different legal systems and regulations. As a general rule, a surveillance measure must have a legal basis, pursue a legitimate aim and be proportionate in relation to that aim in order to be considered lawful. The assessment particularly takes into account factors such as transparency, the scope of the surveillance and the possibility of mapping or profiling the employee. At the same time, Swedish law shows that employers, through the managerial prerogative, have far-reaching authority to direct and distribute work. However, employees’ privacy interests must be given significant weight in the balancing of interests, meaning that a particularly strong employer interest is required for the measure to be considered lawful. Furthermore, the thesis shows that the current fragmented legal framework is in several ways inadequate in relation to rapid technological development. In particular, the increasing use of algorithmic surveillance and management creates legal boundary issues that existing rules are not fully adapted to handle. Although new EU initiatives, such as the AI Act and the Platform Work Directive, strengthen the protection of workers, uncertainties still remain regarding how these rules should be applied in practice. This is especially the case in traditional employment relationships outside the platform work sector. Overall, it can be concluded that the point at which algorithmic surveillance and management become unlawful intrusions into employees’ personal privacy is largely determined through a concrete and nuanced proportionality assessment. This means that similar measures can be allowed in one situation but not in another, making it difficult to set a clear general limit. However, it is clear that there are strong requirements for transparency, that continuous and extensive surveillance is generally not allowed and that certain forms of algorithmic surveillance and management may be accepted, provided that the employer’s interest carries sufficient weight in the balancing of interests.