The integration of Artificial Intelligence (AI) in the management of companies’ human resources is today an unavoidable reality and an opportunity to simplify, accelerate and optimise processes. The dynamism of AI in the employment context cannot, however, be divorced from the growing legal framework on the subject, which has undertaken, among other things, the goal of ensuring information, ethics and transparency in human resource management processes that use AI.    

In 2023, in Portugal, the Decent Work Agenda (“Agenda do Trabalho Digno”) legislative package introduced important information duties into the Portuguese Labour Code regarding the use of algorithms and other AI systems in the employment context, with the aim of preventing the risks of algorithmic discrimination and guaranteeing transparency in their use, in addition to the reservations already stemming from personal data legislation regarding automated decisions.

More specifically, a duty to provide information on AI systems was established, which must be complied with towards the employee, the workers’ commission and trade union representatives, if any. On the other hand, as regards the right to equality at work and in access to employment, the Decent Work Agenda has strengthened the protection of jobseekers and workers, establishing that the existing provisions in this field apply equally to decisions taken based on these systems.

In addition to the obligations arising from the national legal framework, one must also consider those provided for in the AI Act, especially considering that, under the terms of this Regulation, most AI systems currently used by companies in the context of employment relations are classified as high risk, in terms of the risk of injury they present.

Bearing in mind that, in the field of recruitment, for example, AI is now widely used, notably in writing job descriptions, interacting with candidates through chatbots, filtering applications, preparing interview scripts, or analysing and assessing candidates’ responses and behaviour, it is crucial to ensure compliance with the aforementioned legal framework.

The same goes for performance management, where AI has multiple uses, such as organising work, establishing metrics and objectives, identifying development opportunities, designing reward programmes, developing succession plans, or identifying employees eligible for promotion.

It is, however, essential to bear in mind that AI employment related tools are managed by people and, consequently, when algorithms and input data are not rigorously tested or when they are not monitored throughout their useful life, there is a risk of unconsciously perpetuating biased and/or discriminatory tendencies inherent to those who trained the tool, and which impact on the management of human resources.

The paradigmatic case of the AI tool created by the online sales giant to facilitate recruitment, and which generated discriminatory outputs due to the way the algorithm had been programmed, as well as the case of the dismissal of more than 300 of its workers based on lack of productivity exclusively determined by an AI system installed in its warehouses, highlight just some of the risks arising from the unsupervised use of these tools. 

With ethics and transparency being upheld as key issues for AI in human resources management, is your company prepared to face these new challenges?